Meeting Title: Brainforge <> Property-Vista Date: 2024-02-28 Meeting participants: Jp, Joe Mccorkle, Uttam Kumaran
WEBVTT
1 00:01:18.690 ⇒ 00:01:20.060 Uttam Kumaran: Hey, Joe, can you hear me?
2 00:01:25.250 ⇒ 00:01:27.710 Uttam Kumaran: I can’t hear you
3 00:01:28.010 ⇒ 00:01:30.860 Joe McCorkle: alright now I can hear you. Okay.
4 00:01:31.140 ⇒ 00:01:40.199 Joe McCorkle: Yeah. Wasn’t, wasn’t you? It was me. How’s it going? Oh, it’s going good! How are you doing? Well, thank you.
5 00:01:41.110 ⇒ 00:01:51.300 Uttam Kumaran: How’s how’s everything? I I’ve read a little bit about the company? It’s interesting. It was nice seeing the costar partnership. I’ve been working with a company locally that’s
6 00:01:51.490 ⇒ 00:01:56.520 Uttam Kumaran: doing some stuff related to costar data as well, so cool to see that.
7 00:01:57.840 ⇒ 00:02:00.640 Joe McCorkle: Yeah, I mean, yeah, co-star
8 00:02:00.860 ⇒ 00:02:05.719 Joe McCorkle: apartments.com, you know. Couple of other ils. But yeah.
9 00:02:07.820 ⇒ 00:02:08.840 Joe McCorkle: we
10 00:02:10.410 ⇒ 00:02:25.859 Joe McCorkle: help properties get their lease. there’s lease apps completed, and occupancy rates up, and then after that, we help with the tenants, you know, help with maintenance and everything else that they might need
11 00:02:27.360 ⇒ 00:02:29.909 I know what I think we’re
12 00:02:30.470 ⇒ 00:02:33.920 Joe McCorkle: are we waiting on a couple more people? I know.
13 00:02:34.350 ⇒ 00:02:39.460 Uttam Kumaran: jp, also said he’s gonna be responding yes to it. Otherwise we can.
14 00:02:39.540 ⇒ 00:02:42.020 Uttam Kumaran: You know, we can kinda get started happy to give an intro.
15 00:02:42.580 ⇒ 00:02:47.850 Joe McCorkle: Yeah, we can. We can get started. I know Lenny won’t be able to make it. He’s he’s traveling right now.
16 00:02:47.980 ⇒ 00:02:54.190 Uttam Kumaran: Cool. Yeah. I got an intro via Clint, who, I think? You know, Matt.
17 00:02:54.200 ⇒ 00:03:07.409 Uttam Kumaran: part of your team. I think, Lenny, I’m at a conference and me and Clint have worked together and still do some work together. I run an AI and data development firm. Okay, jp, is here.
18 00:03:08.370 ⇒ 00:03:09.949 Uttam Kumaran: hey? Jp, nice to meet you.
19 00:03:10.060 ⇒ 00:03:11.310 JP: Nice to meet you.
20 00:03:11.520 ⇒ 00:03:36.889 Uttam Kumaran: I was just given a brief Intro basically, II got an intro to y’all via Clint. He mentioned that you guys are potentially building some data products. And it’s something that’s kind of in my wheelhouse and done before II currently run brain forage where a data and AI development firm I’m based out of Austin, Texas. But most of the team is is scattered a little all over the place. Have you? Have you visited Austin before?
21 00:03:36.960 ⇒ 00:03:50.139 JP: I? You might not know what this town, but I live in event tech. So that is at the crossroads of 281 and 84. Do you know where that is?
22 00:03:50.140 ⇒ 00:04:14.239 JP: So if you come up. If you’re coming up to anyone, you pass, land passes, and before before you get to Hamilton I’m in the middle of those right, and then. So I’m directly west of Waco hour and 15 min. Yeah. So I go to Cedar Park to go to Costco. That’s I go.
23 00:04:14.280 ⇒ 00:04:14.980 Joe McCorkle: Yeah.
24 00:04:15.060 ⇒ 00:04:35.770 Uttam Kumaran: My career started in New York, and I spent a bunch of time. And then I moved here to Austin a few years ago. I kind of worked as a data engineer, leading data teams and building data products at a couple of different firms.
25 00:04:35.770 ⇒ 00:04:58.490 Uttam Kumaran: And then the last year kind of decided to split off and sort of start a company and go try to help a couple of different firms do things all across data, and building a a team to kind of serve them. So you know, we work with clients doing a lot of internal reporting, doing customer, facing dashboards, embedded dashboards. We handle everything from.
26 00:04:58.540 ⇒ 00:05:26.790 Uttam Kumaran: you know, building doing data, engineering building data models all the way to actually, you know, procuring tooling and actually developing some of the, you know, customer facing dashboards. So kind of like have worked kind of everywhere in in the stack on data side. So basically and that probably most relevant. II had spent a couple of years that we work working really on all the data related to them. So everything from occupancy to finance. I worked on the Ipo
27 00:05:26.990 ⇒ 00:05:31.520 Uttam Kumaran: so kind of really familiar with the commercial real estate business.
28 00:05:31.570 ⇒ 00:05:51.269 Uttam Kumaran: yeah, it’s a bit about me I would love to hear about. Y’all have spent some time reading about the product. And you know, someone familiar with, you know, property management software and that whole world. But you know, definitely, probably missing some nuance, and then, of course, interested to to hear about some of your vision for for data. And you know, data analytics.
29 00:05:53.720 ⇒ 00:05:55.040 JP: you wanna go first, Jim?
30 00:05:56.480 ⇒ 00:05:58.880 Sure. Sure. So.
31 00:05:58.980 ⇒ 00:06:07.070 Joe McCorkle: From a I mean from an AI perspective from a wh. What are we using? Generative AI for today? More for
32 00:06:07.160 ⇒ 00:06:11.039 Joe McCorkle: virtual. Take a virtual assistance.
33 00:06:11.050 ⇒ 00:06:23.469 Joe McCorkle: virtual agents. Whether that be a voice virtual agent. We have one in Beta today. We have a a semi Well, not not really AI
34 00:06:23.580 ⇒ 00:06:26.489 Joe McCorkle: integrated today. But we have a chat bot
35 00:06:26.510 ⇒ 00:06:33.860 Joe McCorkle: which will be integrated soon. So we’re really trying to use generative a I to help people who are shopping
36 00:06:34.010 ⇒ 00:06:50.989 Joe McCorkle: for a place to live. Get information about that. You know, critical pieces of information are things that are in our private data. so not necessarily on the website that can be searched or crawled. You know, it’s it’s private data.
37 00:06:51.030 ⇒ 00:06:57.869 Joe McCorkle: sometimes through the Ils, you can go and scrape that data, get it. But it’s still
38 00:06:58.660 ⇒ 00:07:01.290 Joe McCorkle: They’re considered a little bit private. So
39 00:07:01.590 ⇒ 00:07:05.829 Joe McCorkle: but we step outside of that sometimes. When
40 00:07:06.250 ⇒ 00:07:08.969 Joe McCorkle: a a prospect or tenant may want to know
41 00:07:09.440 ⇒ 00:07:29.219 Joe McCorkle: things about the surroundings of the property restaurants around it. Is there? A place I can pick up my prescriptions. You know. What? What’s the schools? The local, you know, is there kindergarten placenaries or daycare near, is there, you know? Where’s the Close Starbucks gas station. All these things. So you know, geography type base. But then
42 00:07:29.580 ⇒ 00:07:31.450 Joe McCorkle: also things that
43 00:07:31.750 ⇒ 00:07:35.199 Joe McCorkle: genre of AI I could really really help with.
44 00:07:35.320 ⇒ 00:07:56.829 Joe McCorkle: That’s not in our private data. So and everything outside of that. So if you take a a conversation and you build it like, let’s say, with dialogue flow right? You’re going to have intense. So what happens when you don’t have an intent that matches that to me you would go out to generative AI inject some information and the conversation like the address of the property.
45 00:07:57.180 ⇒ 00:08:09.199 Joe McCorkle: and then complete the conversation with whatever the caller asked, or the or the chat person asked. So at least the General AI knows the property, the locale, wherever they add, and then can
46 00:08:09.530 ⇒ 00:08:13.670 Joe McCorkle: can use that injection to answer the question pretty fast.
47 00:08:14.010 ⇒ 00:08:28.820 Uttam Kumaran: Okay, makes sense. And is that that’s primarily just for external. Is there anything like internally, you guys are using generated eye or heavy more on the data analytics side. Currently, I to give you context, one of our customers, one of our clients currently
48 00:08:28.820 ⇒ 00:08:44.810 Uttam Kumaran: they’re planning on implementing some generator, AI solutions on all their Zendesk tickets, and I just read some stuff about how the clariner the company is using a lot of their chat bots, using generative AI to kind of clear a ton of tickets. So there’s a lot of use case I’m seeing on
49 00:08:44.860 ⇒ 00:08:49.679 Uttam Kumaran: like, definitely the service desk customer service side for resolving those.
50 00:08:50.140 ⇒ 00:09:14.269 Joe McCorkle: Ii think that would be helpful to us, but I think even more helpful would be productizing something that looks at things that are repetitive at a property maintenance, request that are repetitive or maintenance request for a certain thing in a certain building that are repetitive. You know things that we can use to say, Hey, this looks like it could be a problem. Throw a flag up for the
51 00:09:14.280 ⇒ 00:09:22.330 Joe McCorkle: property managers, somebody to take a look at, hey? We see this, or you’re trending this way or this looks like to be low. This could be an issue.
52 00:09:23.490 ⇒ 00:09:24.809 Uttam Kumaran: Okay, makes sense.
53 00:09:25.450 ⇒ 00:09:39.220 Uttam Kumaran: And then, how far are you guys currently on your journey with with that product you mentioned, there’s some stuff internally that you guys are using. Generate AI from. There’s some stuff that’s kind of still the normal process of doing help desk or chat up.
54 00:09:39.450 ⇒ 00:09:42.529 Yeah, I mean, from a data perspective, we’re we’re
55 00:09:42.780 ⇒ 00:09:44.030 Joe McCorkle: in. We’re
56 00:09:44.040 ⇒ 00:09:48.239 Joe McCorkle: working real hard to build a total events based architecture.
57 00:09:48.260 ⇒ 00:09:49.700 Joe McCorkle: Okay? So
58 00:09:49.840 ⇒ 00:10:03.759 Joe McCorkle: we’re we’re building events, and those are critical. Those are events to the leasing process. Those are events to signing a lease, getting a lease. Those are events into maintenance. Request service request by tenants.
59 00:10:03.830 ⇒ 00:10:05.089 Joe McCorkle: so we’re we’re
60 00:10:05.660 ⇒ 00:10:15.300 Joe McCorkle: building and growing that events based architecture now. And that’s I think you guys or your team likes
61 00:10:15.440 ⇒ 00:10:19.249 Joe McCorkle: Dvt and Snowflake, we’re consumers of that. Now.
62 00:10:19.300 ⇒ 00:10:30.479 Uttam Kumaran: Cool. Okay, yeah. I mean, I can’t say enough about doing things, and more like an events-based architecture. And you can have like a pretty much they describe as an activity stream, and it helps to do a ton of
63 00:10:30.520 ⇒ 00:10:52.920 Uttam Kumaran: really easy analysis. If you have events tied to objects, and then you can have dimension tables related to that. And then, yeah, you know we’ve me, and you know part part of the couple of people on the team I’ve been doing Dbt for like 5 years now. We we brought on Dbt. At Snowflake at A. We work in 2018, and then I’ve been using Snowflake now, probably like 6 or 7 times I’ve I’ve purchased that
64 00:10:52.920 ⇒ 00:11:12.339 Uttam Kumaran: and then all our current clients are also flake now so very, very familiar with the product. Just actually went to one of their like user group meetups last night. They’re showing off their new AI cortex features. They have like semantic summaries, translations, a lot of interesting stuff. That’s, I think, in preview. So
65 00:11:12.460 ⇒ 00:11:27.990 Uttam Kumaran: yeah, very familiar with with both of those. And so that’s how we’ve been. We’ve been running. You know, we go to a lot of companies. We help them move data, either. Via, you know, custom pipelining or using 5 train to do. Etl, yeah, all of our expertise is in modeling. So whether that’s
66 00:11:27.990 ⇒ 00:11:49.499 Uttam Kumaran: sale data ticket data, finance data. You know, we’ve worked in several of these domains and including a lot of folks from my team, are people II originally met at. We work. So we’ve done all this stuff on the real estate side. And then, you know, making that reportable. So having, like Kpi tables, aggregate tables, anything that your data science team or your an analyst team or your bi team need
67 00:11:49.510 ⇒ 00:11:56.910 Uttam Kumaran: to kind of do their jobs. And then, of course, like anything on the warehouse side. So setting up warehouses, security things like that.
68 00:11:59.240 ⇒ 00:12:03.600 JP: Yeah, we have. So you probably know, we’re multi tenant part of our.
69 00:12:03.910 ⇒ 00:12:06.820 JP: we have a cross so we have
70 00:12:06.900 ⇒ 00:12:10.729 JP: property management companies that manage properties.
71 00:12:10.770 ⇒ 00:12:32.920 JP: But we also have owners who may use multiple property management companies, and they want to see their data right? So as I’m pulling, I can’t give them access to all the Pmc’s data and what what I want to give them access without having to log in and out. Right? So just be across like one of our investment
72 00:12:32.920 ⇒ 00:12:55.570 JP: companies. They have 14 different property management companies managing their properties right and so and there, there, you know, there was like, well, we wanna see this, and we’re almost like, Oh, so people do extract from every property management company. We send them data. Right? So, that’s what’s a challenge there. Another interesting challenge we have.
73 00:12:55.670 ⇒ 00:13:20.179 JP: I would like to this one. So so just I’m gonna give you a quick history, too. So Joe and I joined property vista like 2 wish and a half years ago. Like 2. Yeah, it’ll be 3 years this summer. So we didn’t. We’ve built we? I know we didn’t build. We didn’t build kind of the it. So the product was built over time.
74 00:13:20.290 ⇒ 00:13:23.229 JP: by a group of engineers who were.
75 00:13:24.700 ⇒ 00:13:35.660 JP: it was engineer solving problems in the engineering fashion for engineers, right and so we have. You know, we have some ancient technology.
76 00:13:35.780 ⇒ 00:13:45.060 JP: I joke about like and and and a data model that no no one thought about right, which is.
77 00:13:45.130 ⇒ 00:13:58.420 JP: which is kind of a challenge like even II mean to this day, I’m like, Hey, can I get? Can I get a er d like I just for me. I like to look at those still like from my history. Right? We don’t have one right like
78 00:13:59.020 ⇒ 00:14:02.599 Joe McCorkle: So so we have some challenges there. So the strategy has been
79 00:14:02.970 ⇒ 00:14:10.689 JP: we, which we kind of our strategy of go forward is we building new products kind of on the from an aws
80 00:14:11.460 ⇒ 00:14:25.209 JP: bring the wash, call strategy and plugging them in. And and the other thing is, whenever something happens in one of our systems that just like data that we know people are going to be interested in, we we raise the business event.
81 00:14:25.780 ⇒ 00:14:30.700 JP: So so it’s like, Oh, someone did. Someone opened a maintenance request.
82 00:14:30.790 ⇒ 00:14:34.180 JP: Someone scheduled an inspection. Someone made a payment.
83 00:14:34.220 ⇒ 00:14:57.089 JP: Right? That’s a business event. That business event then goes on a queue, and then it gets consumed and pushed into Snowflake. We may do other stuff with the 2 right, like other systems may need. Oh, that will kick off the survey. So then the survey engine kicks off a survey. But then it that data when it’s done. So the idea is, I can’t afford to re architect what’s there? But I need to have modern
84 00:14:57.200 ⇒ 00:15:15.320 Uttam Kumaran: analytics. Can you talk about the current? So it seems like you do have some sort of like event system or event queue like? Are those all coming just from? I assume the app is installed like a front end, and like a mobile app. And these are all just being triggered. And
85 00:15:15.430 ⇒ 00:15:31.400 Uttam Kumaran: is it the process by which you say, Hey, we want to collect events on this button, or we want to collect events in this flow we’ll add another event. Here’s the properties that goes to the collector. Okay, so there’s some standard. We try to have a kind of a there’s somewhat of a standard payload that we try to add to every event.
86 00:15:31.400 ⇒ 00:15:45.579 JP: right? So that we can type. So again, we can tie that data together in the back. So what’s the unique customer? Id. What’s the need building? Id. Oh, then, is that building associated with one of these owners that might have a access across things like, so,
87 00:15:45.580 ⇒ 00:16:03.020 JP: yeah, so that’s the idea. So when we want one, we have to add one, right? So we have a, we have a queue. So we have some that we finished, and we have a backlog of ones that are being added and we’re trying to roll out some things to customers and get some feedback right? So now we have data, and we can build some dashboards. But it’s like, okay. But
88 00:16:03.350 ⇒ 00:16:24.530 JP: at some point do we have to think about? Oh, how do we get? You know? So it’s analytics. But it’s not insight. So I can be like, Yeah, this is how many you did. And this is how many. And this is how long they took right. And this is the buildings that they’re in. And this is who did them. And this is the types of things. But I don’t have those, Joe said. Oh, wouldn’t it be great to let them know that
89 00:16:24.590 ⇒ 00:16:26.869 JP: in this building, right?
90 00:16:27.160 ⇒ 00:16:30.620 JP: You continue to have the same plumbing problem.
91 00:16:30.970 ⇒ 00:16:38.810 JP: You’ve had it 12 times in the last 6 months, right? Like, I don’t have any insight.
92 00:16:39.570 ⇒ 00:16:40.480 JP: And
93 00:16:40.720 ⇒ 00:16:44.139 JP: and except for maybe the people on this call just
94 00:16:44.380 ⇒ 00:16:45.999 Joe McCorkle: I don’t have
95 00:16:46.230 ⇒ 00:16:55.219 JP: that mindset to to think through that like I have people who can raise the events. I have people who can push the data. Then I can put a dashboard in front of that data, put them like.
96 00:16:55.710 ⇒ 00:17:00.939 JP: but we’re missing the some of the data science, the data engineering that kind of that.
97 00:17:01.130 ⇒ 00:17:06.399 JP: But from this around, oh, well, here’s an obvious Kpi for maintenance. Right? Like.
98 00:17:06.690 ⇒ 00:17:35.829 Uttam Kumaran: right? Yeah. I mean, I know turnaround time. But that’s you know, it’s a very s like, I would say, very common problem, one that I’m dealing with with another client, one that I’ve did particularly I worked at this company called Flow Code. I was leading the data team. There flow code. If you watch super bowl. You saw the circular QR codes. It’s the company that develops that. They’re they were a startup when I joined you know, right before the pandemic, and then, right after I join.
99 00:17:35.830 ⇒ 00:17:59.189 Uttam Kumaran: And and again, like, I got very lucky with that one the QR. Codes just blow up, and you know. We not only developed all the internal reporting, but quickly. We saw that a lot of people were using us for advertising people like Tnt, Esp. And ABC, and then we developed their customer facing data products. But you’re exactly right, is it’s a back and forth. It’s one like, what would we love to show. And then what do we have to show? And then how do we bridge the gap
100 00:17:59.190 ⇒ 00:18:28.310 Uttam Kumaran: over time? We had clients that would come to us and say, we want to see frequency. We want to see where the IP look addresses are. We’re like, Okay, well, we now need an IP lookup vendor. We now need to like structure these things. So it’s a mix. And then you’re totally right, is the transition you need to go through is having all of your data in one place and activity stream right? Having all of that, have dimensionality for your tenants, for property managers have these different objects, and the relationships which are many to one which are related to who. And then, of course, having
101 00:18:28.350 ⇒ 00:18:50.889 Uttam Kumaran: some sort of data, security and understanding what data of each of those roles have access to. And then, second, is at least a minimum being to show historical data. Right? You can show over time what are what are your tickets over time. Who are your what’s your occupancy rate? The next step after that is then, of course, showing like forward looking forecast or backward looking insights showing like, Hey, these are some sort of segmentations that we’ve done
102 00:18:50.890 ⇒ 00:19:09.969 Uttam Kumaran: where we can see that there are ratios, and these are where your your ticket categories are much higher compared to a benchmark. Right? So these are all the classic things that as you guys start sitting on all this data, there’s a huge opportunity to not only monetize that, but make it actionable before just saying, Here’s all your historical data. It’s actually saying, Here’s everything. But also
103 00:19:09.970 ⇒ 00:19:19.609 Uttam Kumaran: you had the same exact as you said, the same maintenance category issue come up 5 times higher than your historical average. So, being able to layer that sort of
104 00:19:20.530 ⇒ 00:19:31.950 Uttam Kumaran: those sort of insights on, and then, of course, layering on forecast so layering on things that are more for facing. And, you know, layering on things like generative AI, where you know, even for your clients, you can put in having
105 00:19:31.950 ⇒ 00:19:54.329 Uttam Kumaran: generally as summarize long ticket requests, or things that are, you know, quite large, and you know. Of course, you guys can monetize and sell that product. So there seems to be a a pretty clear path towards that. So what’s the journey right now in terms of getting that event data into Snowflake. Is that just being processed via like S. 3 and snowpipe? Or what is is that happening today. And if you guys can
106 00:19:54.350 ⇒ 00:19:56.719 Uttam Kumaran: chat a bit about like the state of that
107 00:19:57.540 ⇒ 00:20:00.500 Joe McCorkle: snow pipe
108 00:20:00.520 ⇒ 00:20:07.769 Joe McCorkle: it’s also when we do some things, then dropping it on, and then picking it up and running it through. Dbt.
109 00:20:08.220 ⇒ 00:20:13.860 Uttam Kumaran: and then what is the outcomes right now is it is, are the outcomes just being primarily used for internal reporting.
110 00:20:14.180 ⇒ 00:20:21.350 Uttam Kumaran: Like the current data model, no internal reporting. It’s all all for product.
111 00:20:21.720 ⇒ 00:20:34.450 JP: ye yeah, it. However, W. We do want our. So our internal folks will go in there we’ll give them an admin access so they can look at the same data and produce reports and have
112 00:20:34.600 ⇒ 00:20:38.479 JP: Qbrs with customers like that’ll be a thing.
113 00:20:39.160 ⇒ 00:20:56.429 JP: so yeah. And so I think part of the pain along the way was like, Oh, here’s the event. And then, oh, well, how do we slice by? How do we know who owns it? Right? Oh, yeah, we forgot to put this id on. So then we added this id, right? So now I’ve got data I’ve got like
114 00:20:56.610 ⇒ 00:20:59.419 JP: 8 months of some data, but only have
115 00:20:59.970 ⇒ 00:21:21.879 JP: I only have 2 and a half months of it, with all of the for it to work. And so I was like, Oh, do we go back and look that up? I mean, we we every time we have an event we’re storing it in S. 3. I was like, Oh, cause I learned this lesson at another company like, Oh, we can replay the events right? We can. We could replay all the events, and if we want to change something, but but
116 00:21:21.970 ⇒ 00:21:32.519 JP: we have to still go in there and then take all those ones that we’re missing the Id. Go look them up and add it onto the document and then replay up right? So it’s part of it as I don’t have the bandwidth
117 00:21:32.840 ⇒ 00:21:38.320 JP: right? And those some of that. So. But yeah, it’s all going in there and then.
118 00:21:38.350 ⇒ 00:21:43.220 JP: There is some visualizations that are now being created. On top of
119 00:21:43.390 ⇒ 00:21:46.420 JP: that. We’re we’re actually trying to get customers
120 00:21:46.850 ⇒ 00:21:50.940 JP: next month to take a look at a couple of the like
121 00:21:51.180 ⇒ 00:21:59.890 JP: different data things. We have some maintenance requests and inspections. We’re gonna have some customers preview some of the interfaces. So we can get feedback on
122 00:22:00.320 ⇒ 00:22:14.130 Uttam Kumaran: like, Oh, but yeah, but we wanna green light if it’s this or red light, it’s that. Okay. But instead of us guessing right? So, and is that is that all built on like an embedded bi tool? Or is that all custom in terms of the actual
123 00:22:14.230 ⇒ 00:22:16.290 Uttam Kumaran: like customer facing reporting solution.
124 00:22:17.370 ⇒ 00:22:18.989 JP: Do you want to tell that story?
125 00:22:19.080 ⇒ 00:22:27.230 Joe McCorkle: okay. So. have you heard of propelled data before?
126 00:22:27.590 ⇒ 00:22:29.580 Uttam Kumaran: I’ve not heard of propel? Now?
127 00:22:31.200 ⇒ 00:22:32.999 Joe McCorkle: Okay, well, you’ll
128 00:22:33.950 ⇒ 00:22:47.769 Joe McCorkle: to quickly explain them from tell data. It’s a I would call it still startup, right? It’s a guys that I know who built this propelled data thing, and they were all former twilio resources.
129 00:22:48.140 ⇒ 00:22:53.490 Joe McCorkle: And so they’re really building a middle layer to Snowflake
130 00:22:54.070 ⇒ 00:22:59.390 Joe McCorkle: and kinda in between DBT. And snowflake. And they’re doing a
131 00:22:59.420 ⇒ 00:23:04.489 Joe McCorkle: data model Caching mechanism and building a
132 00:23:05.000 ⇒ 00:23:23.360 Joe McCorkle: essentially a, a, a SQL query engine that caches all of your data for you preloads it so that we can present it snappy fast to our clients. They may go look at, hey? I want to report that 6 months old. Well, we can present that super fast.
133 00:23:23.480 ⇒ 00:23:28.479 Uttam Kumaran: I see. So similar, there’s a company called Cube, that does something similar. Okay?
134 00:23:28.560 ⇒ 00:23:36.170 Uttam Kumaran: Okay, yeah, no, no. Yeah. I’m basically familiar. I know there’s a couple of the people in this space that are kind of doing cause instead of building your
135 00:23:36.330 ⇒ 00:23:40.249 Uttam Kumaran: product on like an analytical. dB, where the queries are super slow.
136 00:23:40.400 ⇒ 00:23:51.949 Uttam Kumaran: have, like a caching layer. And then, of course, in that integrating not only snowflake, but a lot of other data. So that makes sense. They’re just super smart guys. And they’ve helped us out a lot just from a
137 00:23:51.960 ⇒ 00:23:58.679 Joe McCorkle: hey? They’re super smart. You know. We can tell them the data we’re trying to get to. And they
138 00:23:58.790 ⇒ 00:24:01.150 Joe McCorkle: help us solve that problem. Okay?
139 00:24:03.060 ⇒ 00:24:18.900 Uttam Kumaran: And so that the so propell also includes the front facing visuals. It’s it’s not just like a query engine. It’s also like the yeah. I mean, they’ve recently built like some react components and things that we’re we’re using their library.
140 00:24:19.030 ⇒ 00:24:19.720 Joe McCorkle: Yep.
141 00:24:23.950 ⇒ 00:24:29.829 Uttam Kumaran: okay. I mean, that’s I mean that that makes a lot of sense. The one thing that you know, we have a lot of questions from people on are like.
142 00:24:29.990 ⇒ 00:24:43.360 Uttam Kumaran: if they want to build stuff on their own, or if they want to use one of their Vi tools to then do the embedded versions. It’s good that these guys exist, cause I’m sure it’s it’s really quick a lot of people they they try to use like thoughtspot embedded, or something like that. And it’s
143 00:24:43.710 ⇒ 00:24:49.260 Joe McCorkle: I mean, you can. You can use high charts and other things, but you still have to to build the query.
144 00:24:50.310 ⇒ 00:24:53.609 JP: Yup! And then they have some. There’s some also in advance
145 00:24:54.110 ⇒ 00:25:00.790 JP: morning stuff. So our cus our customers could go in. given the right access and write their own queries.
146 00:25:00.810 ⇒ 00:25:08.100 JP: but they can just write their own queries, extract that data, do whatever they want with it. So they can.
147 00:25:08.270 ⇒ 00:25:23.600 JP: Yeah, it’s the advance reporting stuff. Pretty cool cause. Then we can have what’s can. And then we can have this other access for them, because some of our customers have another Vi tool because they’re combining data other systems as well.
148 00:25:23.820 ⇒ 00:25:28.939 JP: and is the is the end product just like a set of dashboards, is it? Also.
149 00:25:28.970 ⇒ 00:25:31.490 Uttam Kumaran: are you guys doing like direct exports like
150 00:25:31.530 ⇒ 00:25:35.920 Uttam Kumaran: setting up pipelines with people? Or is it all through this this interface?
151 00:25:37.660 ⇒ 00:25:45.930 Joe McCorkle: it’s all through the interface. I mean, we build the pipelines. We’re building the pipelines to what our customers need.
152 00:25:46.430 ⇒ 00:25:51.960 Joe McCorkle: you know the only like Chip, you mentioned. Part of the solution will be able to give them
153 00:25:52.490 ⇒ 00:25:58.890 Joe McCorkle: direct, essentially, SQL query access to build their own. Vi reports
154 00:26:01.540 ⇒ 00:26:15.300 Uttam Kumaran: just for their data security, conscious right? Specifically just for their data. So this is where I so one. So on 1 point on the on the actual warehouse side. So one is a lot you can do a lot with Snowflake
155 00:26:15.440 ⇒ 00:26:16.820 Uttam Kumaran: row, level
156 00:26:17.050 ⇒ 00:26:39.520 Uttam Kumaran: access, and like pretty good roles. But again, a lot of that hinges on you understanding the users how if, if, when they’re logging in. It’s like a tenant login. If it’s like an individual user, if there’s different properties within an org like, if there are roles within an org. And then also how that org interacts with all these other problems. So you’re totally right, is like it’s a mishmash of
157 00:26:39.630 ⇒ 00:26:58.040 Uttam Kumaran: objects. And but the nice thing is, Snowflake has a lot of role level security that you can apply and basically at the query level will prevent that data from being, you know, surface, and will apply to every query which is really great. And there’s also really column asking, and a lot of things you can do
158 00:26:58.270 ⇒ 00:27:02.999 Uttam Kumaran: But I would say it seems like the starting point is definitely having some sort of.
159 00:27:03.030 ⇒ 00:27:20.999 Uttam Kumaran: you know, like audit and understanding of all the different objects in the field understanding of, like the schema of like the activity schema, basically, and kind of backing up from like, hey, what’s the vision for this product? What do we have today? What we don’t we have today? And then saying, Okay, let’s do a mapping of, like all of the different
160 00:27:21.210 ⇒ 00:27:37.039 Uttam Kumaran: tenants, property managers, the different roles in between and and kind of come up with. Okay, is there? Is there a scheme by which we can apply these? We can create these roles in Snowflake. But also, basically, do we have all the Ids from each of the properties that we need. And then also, again.
161 00:27:37.040 ⇒ 00:27:51.549 Uttam Kumaran: when people create new events, making sure that your front end team is putting all the right information there and then, also again, having a process where, if the event doesn’t have that, or they can even go validate. Hey? The event is coming into Snowflake
162 00:27:51.970 ⇒ 00:27:59.509 Uttam Kumaran: with all the necessary meta met metadata, or anything you need in a Json Blob, and that you have access to it. That seems like
163 00:28:00.040 ⇒ 00:28:02.079 Uttam Kumaran: the the kind of the largest
164 00:28:02.120 ⇒ 00:28:04.800 Uttam Kumaran: you know thing to tackle. But let me know what you think.
165 00:28:06.940 ⇒ 00:28:12.310 Joe McCorkle: That that would be meaningful. Now, I mean we are
166 00:28:12.440 ⇒ 00:28:21.790 Joe McCorkle: there. There is a potential that will or Sunday separate from Dbt and stuff like that.
167 00:28:22.990 ⇒ 00:28:28.190 Joe McCorkle: So I have to be mindful of that. I think. From a
168 00:28:29.380 ⇒ 00:28:47.520 Joe McCorkle: you know, we’re really trying to drive revenue. So things that are are generative. AI, that can help our clients help our customers. You know that that data predictive analysis of problem analysis, you know, of reoccurring events. Like that stuff we don’t have today, I see.
169 00:28:48.200 ⇒ 00:28:54.560 Uttam Kumaran: So it’s so I think it’s it’s almost like 2 part. Then it’s it’s in addition to security, maybe less
170 00:28:54.990 ⇒ 00:29:03.120 Uttam Kumaran: like hinged on snowflakes like unique properties for that. But basically, how do you structure activities come on? How do you structure like a rough.
171 00:29:03.350 ⇒ 00:29:11.110 Uttam Kumaran: role-based access control, that if you were to plug this into any sort of data warehouse environment that you can kind of enable that. And then, second.
172 00:29:11.610 ⇒ 00:29:41.440 Uttam Kumaran: once we get understanding what data you have on hand for us to kind of be able to say, Okay, here’s like, probably 10 interesting things you could do. Given all the activities you’re coming up with, not only just on basic forecasting, but also are there opportunities where you’re collecting really like unstructured data or long form text that we can use generative AI for so almost like, I would say, 3 different kind of through lines. One is kind of like on the Erd object mapping side, one is really on, like.
173 00:29:41.580 ⇒ 00:29:51.180 Uttam Kumaran: maybe just sequel, based or light. Ml, based like forecast and statistical analysis. And then, third is like, where is there opportunity to plug in.
174 00:29:51.340 ⇒ 00:29:55.449 Uttam Kumaran: you know. Generate AI and and drive like meaningful return on the product?
175 00:29:58.780 ⇒ 00:30:01.439 Joe McCorkle: Yes, yeah, that sounds
176 00:30:02.410 ⇒ 00:30:12.489 Joe McCorkle: that sounds good. It sounds like a lot, but I don’t know that it is right. It’s it’s all starts. It’s kind of like starts one by one.
177 00:30:12.750 ⇒ 00:30:38.650 Uttam Kumaran: I will say, though, like I, even just looking at some of the different products, you guys offer very, very similar, probably data wise to some of the stuff we did we work? Basically, II think that you know, I’m very familiar with kind of like how to model. And people on my team are very familiar how to model that sort of like tenant data, leasing data, and then occupancy data. Think the rest of like tickets and things like that. I think that we’re also have done.
178 00:30:38.710 ⇒ 00:30:47.679 Uttam Kumaran: you know, many times. But let me know, what do you guys think like an engagement? Looks like? I mean, we can put together a little bit more insight from our side about like
179 00:30:47.960 ⇒ 00:30:52.000 Uttam Kumaran: what this could look like. II would need a little bit more context on
180 00:30:52.110 ⇒ 00:31:01.960 Uttam Kumaran: like the the current implementation and and where we would plug in. But what do you guys think is like the most impactful thing if we wanted to try something for a few weeks and
181 00:31:02.110 ⇒ 00:31:08.589 Uttam Kumaran: you know, try to move some move, some move, some move the needle for, you guys.
182 00:31:11.030 ⇒ 00:31:13.890 JP: is it plugging in the generative line, Joe.
183 00:31:15.990 ⇒ 00:31:17.399 JP: I don’t know. I mean
184 00:31:17.480 ⇒ 00:31:22.840 JP: no one’s working on it, right, or is it? Well, I think I think you need to see what we’re building
185 00:31:22.880 ⇒ 00:31:31.910 JP: alright. I think we can do. We can get a fairly quick demo together and just show it to you. Don’t give me access yet, because.
186 00:31:32.120 ⇒ 00:31:47.519 Joe McCorkle: you know, they’re worried. I’ll break it which I will, which I will. Since I touch it, I’ll be like, that’s broken. Yeah, we’ll be getting some
187 00:31:47.700 ⇒ 00:31:57.160 Joe McCorkle: strange shots and some weird. weird notes. But I mean, and also from my point, since from a data science perspective, we have
188 00:31:57.180 ⇒ 00:32:00.220 Joe McCorkle: 13 years of data in our postgres.
189 00:32:00.280 ⇒ 00:32:03.700 Uttam Kumaran: Oh, wow, okay. So we got a lot. Okay.
190 00:32:03.920 ⇒ 00:32:10.270 Joe McCorkle: whether it’s 5 train or whatever. How can we extrapolate that data? Meaningful data
191 00:32:10.650 ⇒ 00:32:23.600 Joe McCorkle: so that I don’t want to go back 13 years. But let’s just say, if we went back a year to get meaningful data out of there that we can then use a I. We can use
192 00:32:23.620 ⇒ 00:32:26.960 Joe McCorkle: analytics tools to give us that data.
193 00:32:27.800 ⇒ 00:32:43.339 Uttam Kumaran: So one thing that would be super helpful is for me to get a sense of the shape. And like, what objects you guys have access to? II heard you know the use case on the main inside. But to even go beyond that, it’d be interesting to hear what are the initial products that you guys are going after. But even just seeing all the different
194 00:32:43.450 ⇒ 00:32:47.970 Uttam Kumaran: objects that you have will allow us to kind of think about. Okay beyond maintenance.
195 00:32:47.990 ⇒ 00:33:00.600 Uttam Kumaran: Be honest with the occupancy data, like, what are some unique data points you guys have, especially if it’s going back that far and then understanding, of course, what’s the first initial product that you’re going after? And how do we? Can we squeeze anything in while
196 00:33:00.650 ⇒ 00:33:02.319 Uttam Kumaran: you know that’s still in development?
197 00:33:05.320 ⇒ 00:33:10.080 JP: We could definitely, I mean, we have a list of all the business events that are either
198 00:33:10.190 ⇒ 00:33:18.389 JP: been created or planned to be created. So I can give you kind of that. We can go. We could definitely go through all that
199 00:33:20.450 ⇒ 00:33:32.489 JP: just. I just had a brain thing where I was like, oh, wait a minute. I was like, if we pass over certain data, we could just run renewal out of there, Joe, for any day of any year of any time.
200 00:33:32.620 ⇒ 00:33:35.789 JP: instead of all this other stuff we’re doing just made me think
201 00:33:35.820 ⇒ 00:33:37.250 sorry about that.
202 00:33:38.280 ⇒ 00:33:41.779 JP: So yeah, we could definitely all those objects, all those
203 00:33:41.830 ⇒ 00:33:48.009 JP: we can definitely take a look at those, and and some of them will start to be combined. So like, you know, we do maintenance request.
204 00:33:48.180 ⇒ 00:34:03.800 JP: But we’re adding these surveys right? So survey will always be related to a maintenance event. But like they’re different business events, right? And we won’t always get a survey resulting. Some people won’t answer a survey, so every maintenance request won’t have one.
205 00:34:04.050 ⇒ 00:34:16.510 JP: but they are always tied to maintenance. So when you do get survey responses, they’re tied to maintenance, which is tied to a maintenance type, right? Like. So it’s like, Oh, whenever there’s a plumbing issue, people are very unhappy.
206 00:34:16.690 ⇒ 00:34:33.599 JP: and they’re very unhappy because and the general feedback is they don’t clean up after themselves right like whatever that is like. We’ll have that data so that there’s kind of like, how do we building those dashboards for those like the insights there like. Oh, you’re happiest. You don’t think
207 00:34:33.730 ⇒ 00:34:50.950 Uttam Kumaran: your staff don’t handle these types of things very well, or these or vendors are not handling these maintenance requests. Well, your feedback is terrible, right? You’re getting those tie back to the tenants themselves, those main events. Okay? So I mean, even understanding, like
208 00:34:51.020 ⇒ 00:34:54.779 Uttam Kumaran: tenant churn probability, or like, you know, looking at
209 00:34:55.030 ⇒ 00:35:12.339 Uttam Kumaran: like impacts to like hey, how many main events are to tenant churn, or at least like, you know, people not renewing their leases or things like that. Yeah, I mean. So even at we work, we did a lot of stuff with the building surveys, with with our like tenant surveys.
210 00:35:12.350 ⇒ 00:35:34.180 Uttam Kumaran: We did a lot on like on the finance side is what I did a lot up towards the end is everything from like, you know economic occupancy, understanding like what you need to do to break even versus like where things are turning and like, what’s your steady state? And you know, we did a ton on turn risk and turn probability and like lead scoring. And so
211 00:35:34.680 ⇒ 00:35:40.550 Uttam Kumaran: again, now, I’m just getting flashbacks to a lot of that. But yeah, sure, it’s very, very. It’s very, very, very similar.
212 00:35:41.400 ⇒ 00:35:46.570 Joe McCorkle: definitely, very similar. And all good data. That’s great data. Yeah.
213 00:35:47.560 ⇒ 00:35:48.900 Joe McCorkle: that we don’t have today.
214 00:35:49.250 ⇒ 00:35:53.650 Uttam Kumaran: But I mean again, if we know that we have all those objects. And
215 00:35:53.790 ⇒ 00:36:11.679 Uttam Kumaran: again, like II live in apartment all II think a lot about even my my my experience. And where the data points are. You know, there’s clear mapping of me submitting events to like. I’m sure there’s a correlation to that versus, you know, renewals, and especially you can tie to the tenant
216 00:36:11.800 ⇒ 00:36:16.539 Uttam Kumaran: and at least provide an understanding of like who’s at risk versus who’s
217 00:36:16.600 ⇒ 00:36:34.979 Uttam Kumaran: I’m already churn and like kind of what the properties of your turn customers were. I think there’s a lot you could do there, but not even that. Going one step further and saying, you should do this right. And I think that’s where there’s probably a lot of use case, for the AI is like you provide it with. Hey, these are the properties of your turn customers. What is the actual
218 00:36:35.080 ⇒ 00:36:37.429 Uttam Kumaran: suggestion you should make?
219 00:36:37.470 ⇒ 00:36:57.910 Uttam Kumaran: That’s the thing where you know. And working with a lot of executives and a lot of operating teams. That’s like the the next level of like, okay, we show historical data. We show some of these like red lights, green lights. But then it’s like, take all that and say, what should I do today, or what should I do this week? What should the priority be? That’s where, again, this is kind of like the cutting edge right now. But that’s where
220 00:36:58.450 ⇒ 00:37:11.950 Uttam Kumaran: we’re trying some stuff with some other clients where we’ve been taking dashboards that we built, passing it right to to open AI image processing and saying like, extract the summary of like today’s data, or like this month’s data. And like, what are the flags?
221 00:37:13.040 ⇒ 00:37:17.290 Uttam Kumaran: So there’s probably some great opportunities linked to just this data product, but probably
222 00:37:17.400 ⇒ 00:37:19.670 JP: otherwise as well.
223 00:37:19.910 ⇒ 00:37:49.070 JP: I guess it’d be super interesting. You know, we one of the products we have is it manages, leads right? It manages the lead to lease process. Right. So we get inquiries from independent listening services. We get inquiries from websites, from the property phone calls, right? And then, we have all the data like, did they? Did they, tour, did they? After the tour? Did they buy, start an application? Did they like, we have all that data?
224 00:37:49.150 ⇒ 00:37:53.709 JP: and we’re just thinking about just simple lead scoring. But it could be that
225 00:37:54.200 ⇒ 00:37:57.760 JP: we could just pass the data
226 00:37:58.920 ⇒ 00:38:04.370 JP: in and then get the get those scores out right? So when we, what we’re trying to do is
227 00:38:04.580 ⇒ 00:38:18.739 JP: make sure that agents are responding right? They’re just not right. So we also have a product where the AI Leasing assistant answers the phone call and has a conversation, right captures data. And then, based on that data.
228 00:38:18.880 ⇒ 00:38:25.489 JP: I would love to send it into AI and have it come out and say, Wait, that’s a that’s a good lead, that’s a that’s a
229 00:38:25.540 ⇒ 00:38:41.899 JP: it. That’s a hot
230 00:38:42.210 ⇒ 00:38:45.450 JP: trying to make sure that the leasing agent spends their time
231 00:38:45.570 ⇒ 00:38:50.269 JP: on a tour or closing the deal. Right? So
232 00:38:51.980 ⇒ 00:39:01.410 Uttam Kumaran: yeah, I mean, I there’s tons of opportunity now. And now I’m also just thinking of a a bunch of things. But it’s great that you guys have. You guys almost have the, you know, the entire funnel, which is which is amazing. It’s one thing I recognize
233 00:39:01.430 ⇒ 00:39:21.859 JP: is like as a data person. I’m like, Wow, they have the entire process like from front to back, not just like one sliver, but basically all the way, all the way to all the way to all the way to all the way to moving out, and that we have the move in inspection. We have the move out inspection. We have the charges right? So it’s can be like.
234 00:39:22.160 ⇒ 00:39:28.049 JP: like, I bet none of our customers know what the average charge is in each building.
235 00:39:28.510 ⇒ 00:39:42.130 JP: right? For on a move out right? And what and what category is is it holes in the wall, is it. you know? Is it? What is it that they charge the most?
236 00:39:42.340 ⇒ 00:39:44.570 Joe McCorkle: Yeah, or just cleaning, you know?
237 00:39:45.240 ⇒ 00:39:49.480 Uttam Kumaran: Yeah, no, that’s great. No, that’s awesome. Yeah, that’s really cool.
238 00:39:50.400 ⇒ 00:40:06.490 Uttam Kumaran: Okay, I think one. Maybe next step. I think that would be great is love to bring on one person on my team. His name is Brian. He’s worked with me very closely. We work on pretty much the other half of the business. He worked on a lot of like
239 00:40:06.670 ⇒ 00:40:12.270 Uttam Kumaran: when you walked into. We work. If you guys have ever been in a we work, there’s the front desk reception. They have like a building health score.
240 00:40:12.290 ⇒ 00:40:21.199 Uttam Kumaran: Basically, that person and their team for that building is their whole life is based on this one dashboard, but says, who’s moving in. Who’s moving out? Who’s coming to tour today?
241 00:40:21.340 ⇒ 00:40:31.580 Uttam Kumaran: Like, what are open tickets there are who’s who’s like at risk for churning. And then we actually scored every single building that we work at. He was like the guy that worked
242 00:40:31.720 ⇒ 00:40:54.050 Uttam Kumaran: on that product that would love to bring him for like a next conversation, and maybe we just go over a little bit about like what data we have access to. And you know, I’m happy to, you know. Send over an nda and make sure that everybody on my team is is under nda for our clients. And I think that might be a good place to start, and then we can kind of at that point
243 00:40:54.050 ⇒ 00:41:09.279 Uttam Kumaran: hopefully give you an estimate of you know what’s the best thing to tackle. Seems like we’re still a little bit pretty broad. Now, looking at what’s the first thing we can tackle? What sort of resources we know. We wanna move on from our side. But definitely, I wanna bring on some
244 00:41:09.570 ⇒ 00:41:15.659 Uttam Kumaran: I want to bring on Brian, who who has my exact same knowledge from the we work side and have us kind of take a look at the data and
245 00:41:15.700 ⇒ 00:41:19.370 Uttam Kumaran: kind of give you some understanding of like, okay, these are some great opportunities.
246 00:41:22.530 ⇒ 00:41:32.479 JP: Yeah. And we could do the little we can show you what we have. And some of the brainstorming, we figure out what makes sense. And yeah, and and
247 00:41:32.750 ⇒ 00:41:53.110 JP: and maybe we can just open up something and just show the data model. And then you could then win. And then, after you know, after you have smelling salts available in case you could be like. And you could be like, Oh, yeah, if you give me 3 years I could. I could give you a urd
248 00:41:54.610 ⇒ 00:41:58.960 JP: so but that would be interesting to know as well. So
249 00:41:59.000 ⇒ 00:42:03.889 Joe McCorkle: now, hold on, we. We do operate our data model with the arm. So
250 00:42:04.030 ⇒ 00:42:06.539 Joe McCorkle: it’s ever evolving and ever changing.
251 00:42:08.610 ⇒ 00:42:12.429 JP: That’s the that’s the that’s the game that’s the game.
252 00:42:12.540 ⇒ 00:42:20.790 Uttam Kumaran: I’m just happy to hear that you guys are on, you know Dvt, and at least hopefully have some sub version control. And like, that’s a really good. This is step in the
253 00:42:20.800 ⇒ 00:42:23.019 Uttam Kumaran: in a good direction. And
254 00:42:23.340 ⇒ 00:42:35.979 Uttam Kumaran: you know, it’s it’s, I think there is a lot of complication in how many different types of objects, but generally like seems manageable. I think the one thing wanna really ensure is that there’s security audit
255 00:42:35.980 ⇒ 00:42:56.029 Uttam Kumaran: when people are accessing the data. And then, second, I think there’s a ton of opportunity for insights. It’s just like narrowing it down like, what are the what are the couple priorities you want to go after? And even how do we measure that like? Do we go after cost centers that highest cost centers we have there, like the sexiest thing. For you know, people who are asking the initial customers like, What’s the
256 00:42:56.120 ⇒ 00:43:02.349 Uttam Kumaran: what’s the biggest target to gonna go after and and build a Poc. And then how do we get that fast enough to put in front of a
257 00:43:02.520 ⇒ 00:43:08.449 Uttam Kumaran: prospective customers say, oh, if I had this like that would change, you know my day to day so.
258 00:43:13.080 ⇒ 00:43:13.830 and
259 00:43:14.280 ⇒ 00:43:20.439 Joe McCorkle: to to be fair, we’re we’re all all of our clients pretty much. I mean, they’re in Canada today.
260 00:43:20.630 ⇒ 00:43:29.320 Joe McCorkle: and we will tell you that they are not as jazzed on this data that the US. Clients are.
261 00:43:29.880 ⇒ 00:43:31.330 Joe McCorkle: We want to bring them there?
262 00:43:31.510 ⇒ 00:43:42.310 Uttam Kumaran: But they’re not today, right? And in what? In what way like, how does that? How does that like manifest like they see. And they’re like this doesn’t have an this won’t have an impact, or what’s the what’s the feedback?
263 00:43:42.640 ⇒ 00:43:46.900 Joe McCorkle: There are a few years behind the Us. On what is important data and
264 00:43:47.510 ⇒ 00:44:04.469 Joe McCorkle: what moves the needle bit well. And, to be fair, a lot of lot of regions like the in Ontario or Toronto area. Their their occupancy rates like they, they don’t have any availability, right? They don’t, they don’t. They don’t. So so in in certain areas of Canada.
265 00:44:04.660 ⇒ 00:44:19.869 JP: You. You don’t do renewals like in the Us. Right? So so I signed my lease for one year, and I’m paying $1,235 a month for one. Better. Let’s just say right. And then at the end of the year.
266 00:44:19.880 ⇒ 00:44:21.929 JP: I can just say, I want to stay.
267 00:44:22.280 ⇒ 00:44:31.660 JP: and then they can. They can increase. There’s a very small amount of rent increase. They can apply. And then you basically go on a month to month, lease
268 00:44:31.690 ⇒ 00:44:39.090 JP: forever very similar to San Francisco, I think in San Francisco
269 00:44:39.510 ⇒ 00:44:49.630 Uttam Kumaran: they they want you to move out. They’re like, Forget double the rent, but also get a get a contract for a year
270 00:44:49.670 ⇒ 00:44:53.430 Uttam Kumaran: whether having all this on instability for a month to month. Yeah, that’s a tough
271 00:44:54.440 ⇒ 00:45:19.189 JP: they can’t raise to market rates until they kick you out. Somebody. Somebody gives their their notice. I think it’s it’s it’s kind of month to month. But you stuff to give like 60 day notice or whatever it is. So they give 60 day notice. And then they’re like, because they’re gonna raise the rent like significantly. And they’re gonna be able to turn that apartment. So they actually some times look there like people prefer like there are people living there, partly.
272 00:45:19.290 ⇒ 00:45:23.760 JP: you know, 10 years, whatever right? I looked at my apartment in Dallas for 7 years.
273 00:45:24.120 ⇒ 00:45:25.290 JP: Right so.
274 00:45:25.370 ⇒ 00:45:41.390 Uttam Kumaran: And then they’re just staying because they don’t want to get hit with the right increase.
275 00:45:42.280 ⇒ 00:45:43.580 Uttam Kumaran: But yeah.
276 00:45:44.270 ⇒ 00:46:05.449 Uttam Kumaran: but that’s tough. I guess if they’re if they’re having lots. But again, there’s I think it’s maybe just going the distance and showing an insight or showing something that’s actually actionable and actually measuring the impact right? Saying that if you were to resolve X amount of tickets, or if you were to, you know, move the needle by X percent. Here the savings and actually showing that versus just.
277 00:46:05.820 ⇒ 00:46:10.610 Uttam Kumaran: you know, II see a lot of customers that would just put line charts and pie charts up, and
278 00:46:10.750 ⇒ 00:46:22.220 Uttam Kumaran: it’s it takes one more step of analysis to kind of show something that’s like Whoa! I didn’t have. I had no idea, or, well, we’re actually almost basically taking the action for you. So
279 00:46:23.980 ⇒ 00:46:25.709 JP: got it.
280 00:46:27.730 ⇒ 00:46:28.730 JP: so
281 00:46:29.060 ⇒ 00:46:38.100 Joe, should we see if we can, how we can get something on the books so that we could get a little demo from noon, and James, and do some follow up.
282 00:46:39.620 ⇒ 00:47:07.290 JP: or should we just go to Austin? Let’s just go. Yeah. I mean, I’d be happy to come up you know your way, or meet somewhere in the middle and say, Hi! It’s it’s rare that I get to work with people here in in like the area. So I’m in the middle, actually in the middle. Joe can fix my wi-fi you can
283 00:47:07.310 ⇒ 00:47:09.049 Joe McCorkle: put together an awesome trip.
284 00:47:10.000 ⇒ 00:47:22.499 JP: but not during South by Southwest daily. Regular traffic is bad enough. I’m not going doing something. So the list. But yeah, we’re working on something. But we could probably get something. I,
285 00:47:22.570 ⇒ 00:47:23.890 JP: James, a member.
286 00:47:24.060 ⇒ 00:47:44.560 Uttam Kumaran: doing some other Demos. Maybe we can even get it put together where they don’t have to do extra one. We could, or whatever is easiest, you know, and we can even follow the questions just wanna make it easy on you guys. And so we can get this is enough of a taste of like what we’re seeing. But you know I generally have a good idea, but let me know. Maybe we can put something in the books, and then, if there’s anything where we can
287 00:47:44.580 ⇒ 00:47:48.660 Uttam Kumaran: come in and be a fly on the wall for a demo or something. We’re happy to do that as well
288 00:47:51.060 ⇒ 00:48:08.370 JP: got it. And I’m gonna end with the beginning in mind since I was a minute and a half late to the meeting. II am get your pronunciation. Do I get Bootam? II was gonna I was gonna be there
289 00:48:08.640 ⇒ 00:48:09.540 JP: excellent
290 00:48:09.750 ⇒ 00:48:17.260 Joe McCorkle: utam. I’m gonna I’m gonna ask you a question here. please. What what is pool parts to go.com
291 00:48:17.510 ⇒ 00:48:21.269 Uttam Kumaran: full parts to go.com is a
292 00:48:21.420 ⇒ 00:48:39.769 Uttam Kumaran: alright. That’s that’s a that’s another email that I have. But, I sorry I just you. I’m using a new calendar to navigate like 5 different email addresses. I have. But it’s another client of ours that that I’m working for. They’re actually a really large. Us-based pool parts e-commerce company.
293 00:48:39.820 ⇒ 00:48:46.620 Uttam Kumaran: So if you guys have pools and you need pool pumps, or you know
294 00:48:46.760 ⇒ 00:49:00.809 JP: I can’t find them. I can’t find anyone to come out here and build a pool, so I’m probably going to do, buy a container pool made out of a shipping container because and cause. Then I’ll just have some. I can get someone to dig a hole out here. That’s a
295 00:49:00.940 ⇒ 00:49:02.140 JP: right? So
296 00:49:03.260 ⇒ 00:49:05.570 JP: that’s what I’m probably gonna do. So
297 00:49:06.750 ⇒ 00:49:08.689 yeah, I have a pool. But
298 00:49:09.230 ⇒ 00:49:18.719 Joe McCorkle: thankfully, it’s running. Well, no problem. You need a pump. You know, these guys do direct to consumer pumps. They’re actually, it’s actually a really really interesting business.
299 00:49:18.890 ⇒ 00:49:24.520 Uttam Kumaran: and I’ve been doing some. Really, they’re they make. They make quite a bit of money. But
300 00:49:24.530 ⇒ 00:49:30.049 Uttam Kumaran: you know, I helped do a lot of this stuff on shipping and shipping rate negotiation. Learn a lot about shipping.
301 00:49:30.120 ⇒ 00:49:47.819 Uttam Kumaran: Kind of a really really cool business, very seasonal business they’re just about to get into like, they’re really, really busy season. So yeah. So that’s exactly what they found. Is that like, Paul Pool guy, you’re held hostage. Because what are you gonna
302 00:49:47.820 ⇒ 00:50:00.759 Uttam Kumaran: say? No, your pool. You want to open your pool next week. They’re gonna pay whatever. And that guy charges a mark up on wherever he gets the parts from. So they have an exclusive deal with black and Decker, where they’re one. They’re one of the only resellers of black and decker.
303 00:50:00.780 ⇒ 00:50:03.010 Uttam Kumaran: you know. Pull pumps and
304 00:50:03.030 ⇒ 00:50:09.089 Uttam Kumaran: they sell like they sell directly to people, and you can install yourself, and it’s really well made. And
305 00:50:09.440 ⇒ 00:50:15.450 Uttam Kumaran: yeah, and they’re saving, come customers, a lot of money, cause these are really really expensive, you know, pumps and things like that. So.
306 00:50:16.000 ⇒ 00:50:41.399 JP: Yeah, if I can get you guys a discount call, let me know. I hope I don’t need it. And if I could pitch my building on on property best to let me know. I don’t know. You know. I don’t know many. I mean, there’s so many of these like.
307 00:50:41.420 ⇒ 00:50:44.760 Uttam Kumaran: you know, sort of like luxury class buildings cup popping up.
308 00:50:45.030 ⇒ 00:50:54.909 Uttam Kumaran: But Austin, real estate market is crazy right now. It’s very, very interesting to watch. Very thankful. I’m don’t own any property, for you know, but
309 00:50:55.180 ⇒ 00:51:00.840 Uttam Kumaran: the supply shock right now is crazy. Especially in like this class department buildings.
310 00:51:01.110 ⇒ 00:51:17.320 Uttam Kumaran: And if you cut, if you’ve come downtown, you’ve you’ve spent any any time in the last 2 years. You see how many cranes are up, and I was just waiting because I. When I renewed my place I talked to them. I said, Hey, you guys notice there’s like 5 other luxury class buildings coming online with each with like couple of 1,000 units like, what do you think is gonna happen? And
311 00:51:17.390 ⇒ 00:51:33.440 Uttam Kumaran: they I don’t think they are really understanding like what’s about to happen. And they offer me concession on my renewal. And then that’s what I knew I was like. Oh, like it’s it’s things are gonna come down a lot. So it’s good, if you know like. Oh, hey, this building’s gonna mix
312 00:51:33.530 ⇒ 00:51:38.749 JP: too much free rent. I’d love to renew with you, but they’re giving me 8 weeks. I don’t
313 00:51:38.880 ⇒ 00:51:41.099 JP: not saying I have to have 8 weeks, but
314 00:51:41.170 ⇒ 00:51:44.330 JP: you know, but in yeah.
315 00:51:44.460 ⇒ 00:52:05.269 JP: and and they’d rather give you the concession and keep the rent at the higher rate, because that looks better on the asset value, right? So so like. So for them like instead of what you like. Oh! Instead of giving me 8 weeks. Why don’t you just make the rent? 22 like whatever you know something else, and they’ll be like. No, no, they’d rather have the rent here, give you the give you the concession, because on the books.
316 00:52:05.270 ⇒ 00:52:14.600 JP: when someone values the building they look at, they don’t look at the concessions. They just look at what is the average rent for the unit. So
317 00:52:14.990 ⇒ 00:52:20.460 JP: yeah. So how long have you been there? I’ve been here since I’ve been here.
318 00:52:41.940 ⇒ 00:52:54.210 JP: Fiberglass filter, they do in a higher end, like, you know, higher value. Yeah. So I found that
319 00:52:54.470 ⇒ 00:53:12.490 JP: When I was in the apartment they would come. Change it all the time if I wanted them to. But it was like not. It wasn’t a good filter, so I just people have no idea that you can just ask and get these things change. I haven’t come as often as I remember, cause I’m like this is in the utility
320 00:53:12.500 ⇒ 00:53:28.689 Uttam Kumaran: bill folks like. Let’s let’s get the let’s get our money’s worth. But II love living, I mean, I lived in New York before, and I would say lifestyle is II really love Austin, you know. Ii
321 00:53:28.790 ⇒ 00:53:56.120 Uttam Kumaran: during the pandemic II bought a truck, and I was just driving around the us, and I came to Austin, and I stayed here for about a month, and I was like, Oh, I gotta move here at some point. Went back to New York, and then that was the last year I had there. I was like, I gotta leave everybody in New York. The problem is, we’ll convince you to stay so right like 2 weeks before I left, and I flew in. I got this apartment, and I didn’t tell anybody. 2 weeks before I left. I was like everybody I’m leaving, and then that way, they don’t have enough time to convince me to stay in New York. So
322 00:53:56.430 ⇒ 00:54:01.150 JP: York unique, you know. But Austin’s kinda fun. So
323 00:54:01.290 ⇒ 00:54:12.149 JP: I need to do. That’s for sure as long as, and if you can walk to it in Austin, it’s much better than if you have to drive to it. So yeah, II live right on Eleventh, right? Next to Franklin’s.
324 00:54:12.940 ⇒ 00:54:26.050 Joe McCorkle: Let me know there’s a lot of other barbecue. That’s I think that’s a brand new. That’s really, really amazing. So if I can schedule us on time, let me know. I’d love to
325 00:54:26.190 ⇒ 00:54:28.609 Uttam Kumaran: love to have some barbecue with, y’all, for sure.
326 00:54:28.790 ⇒ 00:54:29.860 JP: Excellent.
327 00:54:31.290 ⇒ 00:54:33.050 JP: Alright II
328 00:54:33.350 ⇒ 00:54:37.710 JP: I need to run it was great to meet you, and we’ll
329 00:54:37.900 ⇒ 00:54:59.049 Uttam Kumaran: Joe, and I’ll get something so we can get something on the book so you can see some more. You can get your other person on, and we can break and brainstorm like what? Something makes sense to something that makes sense. That’s gonna add value that we can do sooner or later. You guys have you? Wanna send our way? Feel free? And then, yeah, look forward to talking again.
330 00:54:59.310 ⇒ 00:55:00.290 JP: Excellent!
331 00:55:00.510 ⇒ 00:55:02.769 Joe McCorkle: Great! Thank you. I’ll talk to you later.