All Dunkin (Lenni Hub) — Statement of Work (SOW)
Date: 2026-04-29
Version: 0.1 (draft — internal + client review)
Client: All Dunkin, Inc. (or appropriate legal name — confirm)
Author: Brainforge AI
Engagement path: DFS (Data Foundations Sprint) — applied as a product and engineering sprint with fixed duration and milestone acceptance
Internal — remove this block from client-facing PDFs. Sourcing, paths, and Brainforge filing only.
Grill-me (pre-SOW) — decision tree
Problem and goals — resolved from 2026-04-24 discovery and the Lenni Hub reference bundle (
Lenni_Hub_MVP.htmlhub prototype)
- Problem: Community-bank digital lending is disjointed (billboard sites → PDFs); All Dunkin needs a shippable Lenni hub (website → calculator → borrower / lender / bank admin) to validate with banks, not another slide deck.
- Sprint 1 success: A shareable, deployed vertical slice plus alignment to the current Lenni product direction, so All Dunkin can run structured demos and gather bank feedback.
- Measured as: output-first this phase (working demo in an environment), with outcome hooks (funnel metrics placeholders) only where they fit the slice.
Commitments and constraints
- We are not re-committing to a full “100 banks at $500/mo” GTM; that is All Dunkin’s long-term product vision. This SOW is Sprint 1 only (2 weeks).
- Dependencies: NDA and read access to All Dunkin’s existing application source (per call: Doak to provide; Brainforge to send NDA). The shared hub HTML export is a UI reference until that access is granted.
- PII / compliance: Full SOC 2 package and bank exam readiness are out of scope for Sprint 1; we will design segregation hooks and avoid storing production PII in Sprint 1 environments unless the client explicitly approves a test dataset.
Risks
- Scope sprawl across six surface areas (per hub tabs: website, loan calculator, borrower, lender, bank admin, master). Mitigation: one primary user journey for the demo and stubs elsewhere.
- Drift between static HTML and production source. Mitigation: time-boxed reconciliation once All Dunkin’s code is open; if access lags, ship from the reference and list deltas.
Ownership
- Product decisions: All Dunkin (Doak) remain owners; weekly cadence.
- Brainforge: Deliver Sprint 1 milestones and a short “Sprint 2 options” note (not part of this SOW).
Filing: Discovery transcript and working notes used to draft this document live in Brainforge lead workspace (vault); do not print paths on the client copy.
Open items before signature: (1) Legal entity name and billing address. (2) Whether Sprint 1 may use synthetic or anonymized data only. (3) Target hosting for the demo (e.g. All Dunkin’s Railway org vs Brainforge — we recommend client-owned for bank optics). (4) Billing schedule (50/50 vs 100% upfront — Section 16).
1. Executive Summary
The problem: All Dunkin is bringing Lenni, a digital lending hub for community bank relationships, to market. Banks need a path from a public website through loan exploration into borrower and internal (lender and admin) workflows, with a roadmap toward analytics and compliance-grade hosting. The team has a Lenni Hub prototype (multi-tab: Bank Website, Loan Calculator, Borrower Portal, Lender Portal, Bank Admin) that must be productized in code, not only demoed as static assets.
The solution: Brainforge will execute Sprint 1, a 2-week, two-person engineering sprint (Managing Lead and one AI-fluent engineer), to: (a) align implementation to All Dunkin’s source once access is granted, or, if access is delayed, implement the agreed hub slice from the shared reference materials; (b) deploy a demo environment; (c) complete an end-to-end hero path (recommended: loan calculator entry to borrower portal with direct messaging enough for a sales demo, and skeletal admin visibility); (d) document multi-tenant boundaries and Sprint 2 options (integrations, KYC, document vault, full analytics) without committing to them in this SOW.
Expected outcomes for Sprint 1: A clickable, hosted demo the All Dunkin team can show to a design-partner bank; a shared backlog and architecture notes for Lenni; less uncertainty on integration and compliance follow-on.
Guiding heuristics: Reuse the product vision from discovery (manual-to-core in phases, cloud first, fraud and identity as a differentiator in later sprints). Prefer one polished journey over six shallow tabs.
Expected outcome (one sentence): After this sprint, All Dunkin can run a credible, hosted walkthrough of Lenni’s core borrower journey with a clear technical path to Sprint 2.
Detail for client conversations: Sections 2–5 state the two-week demo bar (and what is not required to clear it), IN / STUB / OUT by hub screen and by functionality, and core user personas with suggested talking points for All Dunkin.
Investment: Section 16 — USD 15,000 good-faith rate for follow-on work if awarded (see Section 16).
2. Sprint 1 goal (what Doak can show banks in two weeks)
Primary goal: A hosted URL All Dunkin can walk through with a community bank in a 15–20 minute conversation: enough real interaction on the core journey that the product is believable, not a slide deck. Perfection, full back-office depth, and compliance packaging are explicitly not required to clear that bar for Sprint 1.
Product owner caveats (April 24, 2026 discovery), reflected in this SOW
- The first four experience areas of the hub carry the sales story; bank admin and super-admin work was intentionally not deepened in the reference so far—“they get the point across” without full build-out of every admin behavior. Sprint 1 follows the same idea: show the concept, not every admin control.
- Analytics (funnel, drop-off, management reporting) is valuable but was not agreed as the gating item for a first showable build; both sides emphasized core application feel and calculator ahead of a full analytics module.
- Async direct messaging (not live chat) is part of the borrower experience story.
- Core banking APIs, KYC/ID vendors, and document vault are strategic but are out for a two-week first demo unless reduced to a non-blocking stub (e.g. copy and flow only).
If any item above needs to change for a specific bank meeting, the parties document a one-line change to the Sprint 1 slice in writing (email is fine) before the team re-sequences work.
3. Screens and surfaces (IN / STUB / OUT for Sprint 1)
Use the Lenni Hub structure as the common vocabulary with banks. This table is the default; STUB means: visible screen or placeholder with no or minimal interaction, enough to narrate what comes next.
| Hub area | Sprint 1 | Notes |
|---|---|---|
| Bank website (entry from public site) | IN (at least one credible path in) | Establishes “not a PDF” — e.g. loan product entry → Lenni. |
| Loan calculator | IN | Core gating for “can I afford this / is this the right product?” per discovery. |
| Borrower portal (threads + application path) | IN (hero depth here) | Direct messaging pattern (async), not live chat. Happy path for demo. |
| Lender / processor view (messages + application handoff) | STUB or thin | Enough for Doak to explain how centralized lending or branch staff picks up the thread. Full workflow not required. |
| Bank admin (per-bank product links, disclosures, product setup) | STUB | Reference implied narration-only for Sprint 1 unless time allows read-light (e.g. one product list). Not a full self-serve admin build. |
| Super admin / multi-tenant (rollout across many banks) | OUT of demo depth | Single demo “tenant” or labeling is enough. Scale mechanics are Sprint 2+. |
| Analytics (funnels, drop-off, exec dashboards) | OUT for Sprint 1 (optional log placeholder only) | Explicitly behind core app + calculator in priority for a first bank-facing demo. |
One sentence for banks: “Sprint 1 is the borrower-facing spine and calculator so you can see the experience your customers would get; lender and admin are in the product roadmap — we are showing direction, not the full operating control panel yet.”
4. Functionality (IN / STUB / OUT) for Sprint 1
| Capability | Sprint 1 | Rationale |
|---|---|---|
| Public → calculator → borrower journey (happy path) | IN | What Doak can show end-to-end. |
| Async messages between borrower and bank side (simplified) | IN | Core differentiator vs PDF; not full omnichannel. |
| Application steps / form (enough to continue the story) | IN (depth per timebox) | May be short form; SSN/PII only if agreed and safe in demo. |
| Fraud / selfie / ID integration | OUT | Major sales theme later; not required for “something you can show” in two weeks. |
| Core banking / core API post | OUT | Manual or stub “submitted” state is fine for demo. |
| Secure document upload (full vendor) | OUT | Defer; optional link-out or UI shell if needed for story only. |
| Soft credit | OUT | Defer. |
| Funnel + conversion analytics | OUT (optional single event stub) | Explicitly not the gate for first client-facing demo. |
| Per-bank branding / white-label | STUB | One placeholder brand unless All Dunkin supplies assets in time. |
5. Core user personas and talking points (for All Dunkin)
Use these in bank meetings to show who Lenni is for. Copy can be edited by All Dunkin for tone.
| Persona | Who they are | What they care about | Suggested talking points (Sprint 1) |
|---|---|---|---|
| 1. Prospective borrower | Someone shopping on the bank’s site (especially real estate and commercial ag loan types, per your focus) | Speed, clarity, not driving to a branch for basics; trust | “This is the path from your website to a real application, not a dead-end PDF. Borrowers can explore the loan with a calculator, then continue in one place with direct messages to the bank — not a generic chat window.” |
| 2. Bank marketing / growth | Person responsible for the public site and lead generation | Converting traffic; they often do not see which products drive interest | “The calculator and portal are where digital leads turn into real conversations. Sprint 1 shows the spine; a later phase adds the funnel and conversion views you will want to manage spend and site placement.” |
| 3. Centralized lending / back-office | Team that takes the application after the branch (many banks) | Clean handoff, not re-keying from email; fraud awareness | “The demo shows how a message and application land in one flow. We are not hooking your core in the first two weeks — the point is the front door and data shape; core integration is a staged next step.” |
| 4. Branch / loan officer (relationship) | Human relationship before and after digital | Reassurance that digital helps them, not replaces the officer | “Lenni is the digital front of a relationship model. Nothing here removes the officer; it captures the part of the journey the site is failing on today.” |
| 5. Bank risk / IT / compliance (often secondary in first demo) | Evaluators of fraud and vendor | ID verification, data residency, SOC 2 | “Sprint 1 is intentionally not a full compliance or KYC integration. We are proving client experience first. Identity, documents, and audit packages are on the roadmap in clear phases.” |
| 6. All Dunkin (Doak) / rollout team | You and team rolling Lenni to many banks | Velocity, $500 price point, reusability | “The goal is a productized hub so each new bank is config + deploy, not a new custom build. Sprint 1 gives a credible demo; admin and scale follow.” |
6. Objectives
- O1 — A showable, hosted experience that matches the IN/STUB/OUT tables in Sections 3 and 4.
- O2 — Engineering foundation so the next phase is additive (config, environment, one demo tenant, handoff if built outside All Dunkin’s line).
- O3 — Sprint 2 options in writing: integrations, identity, documents, analytics, compliance, multi-tenant hardening.
7. Expected ROI (sales-facing)
Sprint 1 is not only a demo: Brainforge will help All Dunkin make sound early architecture choices (boundaries, configuration, tenant and data model direction) so that future scale—more banks, more products, and deeper integrations—does not require throwing away the first build.
| Outcome | Current state | Target after Sprint 1 | Value |
|---|---|---|---|
| Sales readiness | “When is it ready?” (per discovery) | Hosted demo of hero journey | Unblocks first bank design partnership |
| Engineering risk | Static or fragmented UI vs one shipped path | Aligned code and one deployed environment | De-risks time to MVP |
| GTM | Pricing vision ($500/mo) without visible product | Tangible Lenni in front of banks | Accelerates feedback loop |
| Architecture for scalability | Build and hosting choices still open or implicit | Explicit decisions and tradeoffs documented (what we’re optimizing for in Sprint 1, what is deferred) plus notes in the handoff package for multi-tenant and integration follow-on | Informs roadmap and reduces costly rework as Lenni moves toward many banks and production-grade features |
8. Scope of work
8.1 In-scope — phase 1 (Sprint 1, 2 weeks)
Engagement structure (DFS, milestone sprint): one 2-week timebox, weekly or twice-weekly check-ins, final demo in week 2.
Workstream 1: product alignment and slice definition (days 1–2)
Context: Reference art includes the Lenni Hub MVP (tabs: Bank Website, Loan Calculator, Borrower Portal, Lender Portal, Bank Admin, and related embedded views). The April 24, 2026 discovery discussion prioritizes the borrower journey, calculator, and messaging (async, not live chat), and funnel visibility over time.
Approach
- Use the client-supplied Lenni Hub reference bundle and the April 24, 2026 discovery session as the source of intent for in-scope vs stubbed work.
- Confirm the Sprint 1 hero path with All Dunkin (recommend: calculator to borrower with a message thread; lender as a thin slice if time permits).
- If All Dunkin’s source is available in time, implement in that code line; otherwise implement in a Brainforge-controlled environment and deliver a defined handoff package so your team can merge or adopt the work without guesswork.
Deliverables
- Slice spec (1–2 pages): which screens are live vs stubbed.
- RACI for who approves product tradeoffs during the 2 weeks.
Workstream 2: build and deploy (days 2–10)
Approach
- Week 1: Scaffold the app, deploy to the agreed cloud target, implement calculator entry and borrower shell per the prototype, and the message experience (in-app or same-session relay as in the reference, simplified if needed).
- Week 2: Harden the happy path and error states for demo. Add bank admin as read-light or config (e.g. list products and links) if time allows; otherwise it moves to Sprint 2. Analytics: optional event or step logging in Sprint 1 (placeholder only).
- AI engineering: AI-assisted build workflows under the Managing Lead for speed and quality (standard Brainforge operating model, not a separate line item).
Deliverables
- Deployed URL(s) and a run and deploy document (local run, deployment, and configuration; no real secrets in written materials).
- Demo script (15–20 minutes) for bank-facing walkthroughs.
- Sprint 2 backlog note: integrations (core, identity providers, document vault), SOC 2 program, further multi-tenant hardening.
8.2 Required deliverables summary
| Deliverable | When |
|---|---|
| Slice spec and hero-path agreement | End of day 2 |
| Hosted demo (hero path) and run/deploy documentation | End of week 2 |
| Sprint 2 options and backlog (written) | End of week 2 |
| 30 min demo and walkthrough to All Dunkin | Last day of sprint |
8.3 Out-of-scope — what this engagement does not include
- SOC 2 certification, bank exam support, or live production PII (unless added by written change).
- Direct API integration to core banking systems (later phase per discovery; not Sprint 1).
- Full analytics product in the “complete funnel dashboard” sense — at most instrumentation placeholders in Sprint 1.
- KYC, facial match, and soft-credit as finished product features — vendor selection and wiring in Sprint 2 or later, unless a time-boxed stub is agreed in writing.
- Multi-tenant production at scale; Sprint 1 may use one demo tenant and documented data boundaries.
- Ongoing support, SLAs, or work beyond the 2-week sprint (requires a new SOW or addendum).
9. Requirements and inputs (dependencies)
All Dunkin
- Execute the NDA and, as applicable, the MSA for professional services.
- A point of contact with authority: Doak (with Clarence Stone / Trust Vicinity for coordination as needed).
- Source access and materials as soon as the NDA allows, including the Lenni Hub reference the parties have already shared.
- Branding and copy for the public demo, or written permission to use a neutral placeholder bank.
- No live borrower PII in Sprint 1 without prior written approval; use sample data.
Brainforge
- Managing lead: Uttam Kumaran — scoping, alignment, acceptance.
- Engineer: one AI engineering lead (level TBD) — build, test, deployment.
10. Project timeline
| Phase / milestone | Calendar | Focus | Deliverables |
|---|---|---|---|
| Sprint 1 | 2 consecutive weeks (start TBD) | Hero path, deploy, handoff | Demo link, code handoff package, Sprint 2 options doc |
Total duration: 2 weeks (10 business days, adjusted for holidays with the client if needed).
11. Assumptions
- All Dunkin can hold at least 2 live working sessions in the sprint (kickoff and mid- or end-review) and respond async within 1 business day for blockers.
- One primary hero path is the bar for Sprint 1; other hub areas stay stubs unless the parties agree in writing to re-sequence.
- No separate regulatory sign-off is required to host synthetic data in the agreed hosting.
- English-only product copy in Sprint 1 unless the parties agree otherwise.
12. Risks and mitigations
| Risk | Impact | Mitigation |
|---|---|---|
| Source not available until late in the sprint | High | Start from the Lenni Hub reference; time-box bring-up into All Dunkin’s line when access opens |
| Scope push to full admin, analytics, and integrations | High | 2-week fixed scope; any expansion needs a written change order |
| Third-party (identity, documents) lead times | Med | Sprint 2+; optional stub in Sprint 1 if agreed |
| Hosting and account ownership unclear | Med | Prefer an All Dunkin–owned cloud for bank optics; Brainforge account only if both sides agree in writing |
13. Acceptance criteria and success indicators (DFS)
- A1: A public or access-controlled URL runs the agreed hero path with no manual “developer in the room” to complete the story.
- A2: The agreed primary flows pass a pre-written happy-path script (agreed at mid-sprint).
- A3: The run/deploy document explains local run, deployment, and where configuration lives (placeholders for secrets, not live secrets in shared documentation).
- A4: A Sprint 2 options document with 5+ candidate work items and stated dependencies (e.g. core, KYC, document handling, compliance).
14. Communication plan
- Kickoff (60 min): day 1 — hero path, access, and hosting.
- Sync (30 min): mid-sprint (e.g. day 5) — stubs vs depth.
- Demo/acceptance (30–60 min): last day of the sprint.
- Async: one agreed channel; 1 business day for normal replies.
- Escalation: timeline risk to the managing lead within 1 business day of awareness.
15. Team
Brainforge
| Role | Name | Responsibility |
|---|---|---|
| Managing lead / client alignment | Uttam Kumaran | Product slice, sign-off, demos |
| AI engineer (×1) | TBD | Build, test, deploy |
All Dunkin (client)
| Role | Name | Involvement |
|---|---|---|
| Product owner | Doak | Prioritization, acceptance |
| Facilitator / intro | Clarence Stone (Trust Vicinity) | As needed (not a signatory unless agreed) |
16. Pricing and payment (draft)
Model: Option B — sprint-based (2-week sprint)
- 2-week Sprint 1 (this SOW, fixed): USD $20,000 (includes Managing Lead and 1 engineer as in Section 8).
- Not in fee: pass-through cloud costs and SaaS (hosting, identity, etc. — on All Dunkin’s books or by separate estimate).
Follow-on work (good faith, long-term partnership): If All Dunkin awards Brainforge a follow-on statement of work for continued Lenni work (e.g. Sprint 2 or the next agreed phase), Brainforge intends to offer that follow-on engagement at USD $15,000 for a comparable 2-week sprint scope (or the fee stated in the follow-on SOW if the scope differs), as an indicator of good faith and a long-term partnership. Only the signed follow-on SOW is binding; this paragraph states commercial intent, not a commitment to begin work without a separate agreement.
Billing (propose): 50% on kickoff, 50% on acceptance; or 100% upfront — to agree.
17. Sign-off
By signing below, both parties acknowledge the scope, deliverables, and timeline in this SOW for Sprint 1 only.
All Dunkin, Inc.:
Name: ___________________________
Title: ___________________________
Date: ___________________________
Signature: _______________________
Brainforge AI, Inc.:
Name: ___________________________
Title: ___________________________
Date: ___________________________
Signature: _______________________
Optional: Next steps
- Confirm legal name, data rules, hosting owner, and billing schedule.
- Sign NDA and this SOW when ready.
- Pre–kickoff: access and start date.
- Day 1: kickoff session.