Business Model
📐 SCOPE · WHAT THIS CHAPTER COVERS
The 5 revenue lines. How money enters Nova AI Ventures. What each line is, how it makes money, and when it applies.
§04 answers WHAT we sell and HOW money enters. The math (MRR targets, timing, activation triggers) lives separately in §11 Revenue Generation — this chapter keeps the model clean.
1. Five Revenue Lines
💰 REVENUE LINES · 5 SOURCES
L01 Platform
Nova AI Stack license at buyout
L02 Apps
Nova AI-initiated & co-founded apps
L03 SMB
Productized template deployments
L04 AI Mgmt-Suite
Deployable AI team for clients
L05 Venture R&D
Paid engagements → platform assets
Line 01 — Platform & Tools
the Accelerator- What it is: Nova AI Stack = Nova AI Ventures’ proprietary setup (Layers 01-06 per §06). We provide it to Child SPVs as a one-off license — accelerator kickstart. Improvements from SPVs flow back to the platform. The platform self-improves from portfolio experience.
- How money enters: Only at buyout event. When someone buys a Child SPV, they pay a one-off license fee to Nova AI Ventures. Until that moment: 0 PLN recurring.
- License structure: can be loan · deferred payment · lump sum at exit · Mode 1-4 maintenance fee (§06 E) — Mikołaj + Brian choose per SPV
- Operator requirement: N/A
- Timeline: Value accumulates throughout the entire life of the SPV, monetized at exit
- Key insight: This is not a revenue line in the classic sense. Value-capture mechanism at exit. Does not enter MRR.
Line 02a — Apps
Nova AI Ventures-initiated- What it is: Apps (B2B, B2C) where the idea originates from us. Built on Nova AI Stack + Product Builder. Run as SPV (virtual → formal).
- Origination: Nova AI Ventures founder observes market/problem → Product Builder gates the idea → passing → build
- How money enters:
- Virtual SPV phase: net profit flows directly to Nova AI Ventures (Nova AI Ventures = registered entity)
- Formal SPV phase: SPV runs own books → after cost deduction → split per equity % → Nova AI Ventures gets its share (typically 25%)
- Token metering margin from AI usage in the app (ongoing)
- Exit: equity stake realized at sale
- Operator requirement (hard day-1 protocol):
- Internal responsible person — always from day 1, even for internal apps
- External operator search — always active from start of build (or as early as possible)
- Role: PM · project mgmt · sales · growth driver — “someone who pushes this mega-hard forward”
- Examples: SnapSell (live, operator search active) · Albik Creator / My Story Machine (operator case study) · Iris AI
Line 02b — Apps
Co-founded with industry expert- What it is: Apps where the idea originates from a domain expert — a person with decades of expertise in a narrow field (wealth mgmt, legal, marketing, medicine, education, etc.), with a network and market position, who wants to build an AI-native business on top of their expertise.
- Origination: External expert → outreach or inbound → scoring fit → co-founding partnership
- Value exchange:
- Nova AI Ventures gives: AI technology stack · Product Builder · legal framework (SPV + operator equity structures) · tools (agent orchestration, token economy) · hosting + ops
- Expert gives: domain expertise · network · industry credibility · market position · often first customers
- Operator dynamics: The expert is often the natural operator / co-CEO (because they have domain + network). Or an external operator is hired under the expert’s guidance. Each case structured individually.
- Equity structure: Nova AI Ventures + Expert + (optional external operator) — split differently than 02a, because the expert contribution is not just “idea” — it is substantial ongoing value
- How money enters: Same mechanics as 02a (virtual → formal SPV · net profit flows · token margin · equity at exit) — but equity split differs
- Examples: Lumen Wealth Management (co-founded with Ross Smith · domain expert in wealth management · idea originator + product owner + investor) · Otherth. Marketing Platform (co-founded with partner → future CEO)
Line 03 — Productized SMB Implementation
Reusable Template × Deployment- What it is: Template-driven system with reusable components. Fast deployment + maintenance + security as the offering. Opens a new type of client: SMB (mid/small, sometimes large) that wants quick deployment, often vibe-coded, but does NOT want to think about maintenance and security.
- Value prop: “Make software cheaper, deploy it, secure it, maintain it” — a game changer for clients who have everything in a basic way
- How money enters:
- Setup fee (one-off) — covers Nova AI Stack license for their deployment + initial integration
- Annual subscription (flat, per year; not monthly)
- Token metering margin — ongoing AI usage
- Buyout lump sum — if client wants to take full ownership of the solution
- Token economy layer in L03: Beyond the annual subscription, each SMB deployment runs on its dedicated proxy + server infrastructure (part of Nova AI Stack). The proxy meters all AI usage. Nova AI Ventures earns token metering margin on every AI operation the client performs. This opens a scalable revenue line: we can build simple, token-driven AI tools that solve specific SMB problems (e.g., document processing, CRM automation, report generation) and deploy them as internal products on the SMB line. Each tool becomes a reusable template — built once, deployed many times — with revenue flowing from both the annual subscription AND the token consumption underneath.
- Operator requirement: N/A (template owned by Nova AI Ventures)
- Infrastructure: dedicated server per client (separate MCP instance, separate env, separate billing)
- Example: CNC RadFres — company manufactures parts, has everything in a basic way. Building them a product + hosting + maintenance = game changer.
Line 04 — AI Management Suite
Productized AI Team Deployment- What it is: Deployable version of our own AI Management Team (§07) — a set of “process/procedure managers” for client companies. The client gets a virtual team (marketing, finance, tech, legal, etc.) operating just like our internal suite.
- Mechanism: Dedicated server with full AI Management Team + central information management system + software integration layer (connecting software + knowledge about the client’s business).
- Capability examples: creating presentations · searching for a person on LinkedIn · writing emails · managing HR/finance/marketing processes · reporting
- Interface: Slack, Telegram, or other easy-to-communicate-with chatbot-type interface.
- How money enters:
- Onboarding fee — deployment + analysis of the client company’s business case (upfront)
- Monthly / annual subscription — maintenance + improvements
- Token metering margin — ongoing AI usage
- Token economy layer in L04: Base subscription (typical reference: ~$200/month per client, comparable to ChatGPT Team or similar) covers platform access + included token bundle. Beyond the bundle, clients consume tokens through a dedicated LiteLLM proxy instance running on their deployment. The proxy meters every AI call, Nova AI Ventures applies margin (30-50% pass-through), and the client pays cost + margin. This is the primary revenue layer once the client engages deeply — subscription is the access gate, token margin is where Nova AI Ventures earns compounding revenue.
- Operator requirement: N/A (productized)
- Cross-ref: Internal AI Management Suite described in §07; here the commercial/productized version.
- Anti-pattern: we do not sell one-shot AI consulting. Line 04 ALWAYS has an ongoing subscription.
Line 05 — Venture R&D
Paid external engagement → reusable platform asset- What it is: External engagement funding development of reusable platform assets. Alt framing: “war field playground” — a paid proving ground for building tools/patterns that then flow back to the platform.
- How money enters: Project fees (30-80k PLN per engagement)
- Unit economics: ~25k PLN/month avg retainer
- Operator requirement: N/A
- Rules (hard):
- Max 2 concurrent engagements at any time
- MUST pass Venture R&D Filter (§03 section 5): Extract · Validate · Deploy · Plant (at least 1 of 4)
- Pure bespoke client code = fail = NO
- Reusable asset deliverable required
- Example: Softiq (active enterprise implementation) · RadFres SMB+AI Management (pipeline · may productize into L03 template post-delivery)
2. Revenue Mechanisms Grid
⚙️ MECHANISMS · HOW REVENUE COMPOUNDS
Not every line maps to a single mechanism. The grid shows which mechanism handles which line:
| Mechanism | L01 Platform | L02a Apps · Nova AI Ventures | L02b Apps · Co-founded | L03 SMB | L04 AI Mgmt Suite | L05 R&D |
|---|---|---|---|---|---|---|
| One-off Platform License (at buyout) | ✓ | |||||
| Setup / Onboarding fee (upfront) | ✓ | ✓ | ||||
| Subscription (monthly or annual) | ✓ | ✓ | ||||
| Token metering margin (ongoing) | ✓ | ✓ | ✓ | ✓ | ||
| Net profit flow (virtual SPV) | ✓ | ✓ | ||||
| Equity-proportional profit share (formal SPV) | ✓ | ✓ | ||||
| Project fee (lumpy) | ✓ | |||||
| Equity stake (realized at exit) | ✓ | ✓ | ✓* | |||
| Buyout lump sum (exit event) | ✓ | ✓ | ✓ | ✓ |
(*L05 equity only when it passes the “plant equity seed” test from the Venture R&D Filter)
🧠 MENTAL MODEL · WHAT WE SELL
Mental model per line
| Line | Mechanical shape |
|---|---|
| L01 | Pure exit-capture (buyout only) |
| L02a/b | Ongoing profit/token + exit equity (origination differs; mechanics identical) |
| L03 | Setup + annual sub + tokens + optional buyout |
| L04 | Onboarding + subscription + tokens (SaaS-like) |
| L05 | Project fees + optional equity seed |
3. Anti-patterns · Explicit Rejections
🚫 ANTI-PATTERNS · AVOID THESE
Rules carried forward from §03 Non-Negotiables + Values, applied specifically to revenue decisions:
Anti-pattern 01: Products without operator-search-in-motion — enforced by every new L02 build = parallel operator search from day 1 (§09 Operator Model)
Anti-pattern 02: Consulting for consulting’s sake — every L05 engagement MUST pass Venture R&D Filter (§03.5)
Anti-pattern 03: Bespoke client code without reusable asset — L05 hard rule, fail filter = NO
Anti-pattern 04: L04 deployments without an ongoing subscription — L04 ALWAYS has an ongoing sub, no one-shot deployments
Anti-pattern 05: Products-for-products-revenue without operator path — L02 requires operator day-1 protocol, without it = founder homework
Anti-pattern 06: SMB bespoke (client-specific code that doesn’t templatize) — L03 rule, reusable template required (Non-Negotiable #7)
Anti-pattern 07: Platform license in a way that Nova AI Ventures loses proprietary control — L01 one-off + buy-back clause (§06 E)