CTA DataOps Planning

Date: February 9, 2026 Source: Granola Meeting ID: 72dcde3b-47b8-4e35-bcc1-18674e52c8ee URL: https://notes.granola.ai/t/72dcde3b-47b8-4e35-bcc1-18674e52c8ee

Participants:

Summary

Committee data view implementation ready
- Bridge table needed to connect individuals to organizations
- Display current committee members on company pages
- Query logic sufficient, no structural changes required
New Snowflake role configurations needed:
1. Prod-read role hiding staging/dev databases (show only prodmart and raw)
2. Role for mart and archive access only
3. Streamlit app creation privileges without other write access
Membership team early adopters get Streamlit building capability
- Built audience profiler using Cortex in Snowflake
- Promise: “can’t break anything in Snowflake”

Current challenges with identity resolution:
- Multiple entities share same website domains
- Gmail users with LinkedIn/YouTube as websites
- Samsung subsidiaries inflate data volume when matching on website
Waterfall matching logic priority:
1. Impacts ID or exhibitor ID (highest priority)
2. Company name matching
3. Email domain matching
4. Website domain (rolls up to parent company)
Data freshness hierarchy needed
- Sales team exhibitor links (2027 CES) prioritized over old remembers data
- Formal ID matches trump domain/name matching
Individual name matching rejected as too unreliable
Metrics tracking planned for match success rates
Total unique companies estimated ~4,000 from 4M company name variations

Member engagement tracker demo shared
- S3 logo integration working
- Primary rep job titles added to model
- CES data 2023-2026 included
- Innovation awards tracking
- Virtual event attendance alerting requested
Registration consolidation priority for Kyle
- 2.5M rows in registration history
- Modern history focus for new RPT
Awaish and Ashwini pairing on identity stitching implementation
Working sessions planned to surface edge cases
Daily follow-ups on access provisioning