Gnesis Spec · Go To Market

Gnesis Battlecard

Version 0.2 (directional, not final) Last updated: 2026-04-23 Owners: Rabih Naja (Founder & Chief Builder), Wassim Moumneh (Co-Founder & Chief Orchestrator) Status: Source of truth for competitive positioning


0. How to use this document

This is a sales, investor, and recruiting battlecard. It tells anyone representing Gnesis what Gnesis is, what it is not, how to frame it against competitors by name, which fights to pick, which fights to skip, and what to say when someone pushes back.

Three rules:

  1. Never concede the category. Gnesis is a conversational AI business factory. It is not an "AI app builder," an "AI coding assistant," a "no-code tool," or a "legal template site." If someone bins it into those categories, pull the conversation back.
  2. Never fabricate. Every number in this doc is either a Gnesis-sourced commitment (see "Gnesis-sourced" tags) or a competitor number verified from a public source (see Sources at the end).
  3. Never oversell. The "When we lose" and "Disqualification" sections exist because walking away from the wrong customer is a win.

1. Gnesis in one sentence

Gnesis is a conversational AI business factory. A founder, operator, or team talks to Shakeeb, and Shakeeb produces a complete, multi-tenant, compliance-ready, production SaaS in 72 hours plus 30-plus business-setup documents (legal, compliance, finance, marketing, content, ops, sales) across the following 7 to 14 days. Not a prototype. Not a demo. A running business with software, documents, and trust contract baked in.

2. The category reframe

Most of the market is selling one of four things:

  1. Prototype makers (Bolt, Lovable, v0). Great for screenshots. Break when a real customer shows up.
  2. Developer accelerators (Cursor, Claude Code, Replit Agent). Assume a developer in the seat. Gnesis does not.
  3. Design-to-code platforms (Webflow, Framer). Marketing sites and light CMS. Not apps.
  4. Internal tool builders (Retool). Glue on top of existing databases. Not net-new SaaS.

Gnesis is the fifth thing: a factory that turns a conversation into a shippable product with all the non-negotiables already wired in.

3. The three pillars (never off-message)

Pillar 1. Shakeeb, a voice-first cofounder

Shakeeb is the orb, always present, always in context. Voice round-trip target: 800ms p95 (Gnesis-sourced, Shakeeb spec). Six decision classes: Answer, Ask, Propose, Push back, Gate, Defer. Shakeeb pushes back when he disagrees; he does not rubber-stamp.

Pillar 2. Aleph, the operating system underneath

Aleph is the substrate nobody else has. 13 phases, 352+ skills, 38 workflows, 17 enforcement gates (Gnesis-sourced, Aleph positioning). Every app Gnesis produces runs through the same factory. That is why Gnesis outputs are consistent, not artisanal.

Pillar 3. Ten non-negotiables in every generated app

Every app ships with auth, billing, internationalisation, accessibility, observability, multi-tenancy, compliance posture, security, operational UX, and design coherence (Gnesis-sourced, Generated App Standard). Other tools call these "features you can add later." Gnesis calls them "the floor."

4. Category map

Category Players What they do Where they stop
Prototype factories Bolt, Lovable, v0 One-shot UI or single-file apps from a prompt No multi-tenancy, no compliance posture, no production-grade isolation
Dev-in-the-loop assistants Cursor, Claude Code, Replit Agent Accelerate a developer's IDE work Require a developer, an architecture, a codebase
Design-to-code Webflow, Framer Marketing sites, light CMS Not an app platform, no auth or multi-tenant business logic
Internal tools Retool, Appsmith Dashboards over existing data Not net-new product, not customer-facing
Gnesis Gnesis Voice-first, produces production multi-tenant apps from conversation in 72 hours Not for prototypes, not for marketing-only sites, not for pure internal tools

5. Side-by-side feature matrix

Legend. ✓ = shipped. △ = partial or via add-ons / plugins. ✗ = not a product goal. ? = unverified.

Capability Gnesis Bolt Lovable v0 Replit Agent Cursor Claude Code Webflow Retool
Voice-first interface
Conversational, persistent cofounder
72-hour brief-to-production target ✓ (Gnesis-sourced)
Multi-tenant built in
Per-tenant Postgres isolation mode ✓ (Gnesis-sourced)
VPC-Vault per-tenant isolation ✓ (Gnesis-sourced)
Auth, billing, i18n, a11y baked in by default
Compliance roadmap (SOC 2) attached to product ✓ (Gnesis-sourced) ? ? △ (Vercel) ?
Enforcement gates in CI ✓ (6 gates, Gnesis-sourced) n/a n/a
Model failover across providers ✓ (Anthropic → OpenAI → Google) ? ? ? n/a n/a
Voice p95 round-trip ≤800ms ✓ (Gnesis-sourced target) n/a n/a n/a n/a n/a n/a n/a n/a
Arabic and Spanish default on all tiers ✓ (Gnesis-sourced)
Trust contract with approval gates ✓ (Gnesis-sourced) n/a n/a
Entry price $19/seat + AI pass-through (+20%); typical all-in $40-50/seat/mo (Gnesis-sourced) $25/mo $25/mo $20/mo $20 credits $20/mo $100/mo min for Claude Code $39/site/mo $10/standard user/mo

Unverified cells marked "?" are not to be cited externally without confirmation. Pricing figures for competitors are from April 2026 public pages; see Sources.

6. Head-to-head cards

Each card follows the same skeleton. Keep the skeleton; change the content only.

Card 6.1. Gnesis vs Bolt.new

What they are. A browser-based, StackBlitz-backed prompt-to-app tool. You prompt, it writes a full-stack app that runs in a browser sandbox. Their pricing (April 2026). Free with 1M tokens/month and 300K daily cap. Pro $25/mo with 10M tokens. Teams $30/member/month (credits not pooled). Enterprise custom. What they do well. Instant start, very fast first paint, good for a one-night prototype or a hackathon demo. Where they fall short for our ICP.

Our winning play. "Bolt is a sketchpad. Gnesis is a factory. The moment your prototype gets a paying customer, Bolt's model falls over because it has no tenancy, no billing, no compliance path. Gnesis ships with all of that on day one." Lead with the 10 non-negotiables. What we concede. If the buyer truly just wants a one-off prototype for a pitch tomorrow, Bolt is fine and cheap. Do not fight that fight.

Card 6.2. Gnesis vs Lovable

What they are. A credit-based app builder that turns prompts into working React apps with Supabase on the back. Strong with design-minded founders. Their pricing (April 2026). Free with 5 daily credits (30/month cap). Pro $25/mo. Business $50/mo adds SSO, templates, restricted projects. Enterprise custom. What they do well. Pretty, working outputs quickly. Good onboarding. Supabase-native, so auth and DB are present out of the box. Where they fall short for our ICP.

Our winning play. "Lovable builds one app at a time. Gnesis builds a product. If your product needs multiple tenants, per-tenant data, or regional separation, Lovable makes you solve that yourself. Gnesis hands it to you." Show the three isolation modes. What we concede. Lovable's aesthetic defaults are strong. For a designer-founder building one app one time, Lovable feels great.

Card 6.3. Gnesis vs v0 (Vercel)

What they are. Vercel's generative UI and component tool, now a full app builder with tiered AI models (Mini, Pro, Max) and a $-denominated credit system. Their pricing (April 2026). Free $5 credits/month. Premium $20/mo with $20 credits. Team $30/user/month with $30 credits each. Business $100/user/month. Enterprise custom. What they do well. Deepest integration with Vercel's deployment stack. Good UI output. Figma imports on Premium+. Where they fall short for our ICP.

Our winning play. "v0 is Vercel's way to sell more Vercel. Gnesis is infra-agnostic by design. We use OpenRouter with a failover chain across Anthropic, OpenAI, and Google so you are never stuck." If the buyer cares about data residency or model-provider risk, press hard. What we concede. If the buyer is already all-in on Vercel and wants tighter integration, v0 has a short-term advantage.

Card 6.4. Gnesis vs Replit Agent

What they are. Replit's agentic builder layered on top of Replit's cloud IDE. Long-running autonomous builds via "checkpoints." Their pricing (April 2026). Starter free (limited). Core $20-25/mo with $20 usage credits. Pro $100/mo for up to 15 builders pooled. Enterprise custom. What they do well. Real cloud IDE behind it. Can run long autonomous builds. Strong for developers who want to tinker. Where they fall short for our ICP.

Our winning play. "Replit is for developers who want an agent. Gnesis is for founders who need a cofounder. Different product, different customer." If the buyer is non-technical or operator-first, this fight is easy. What we concede. For a solo engineer who wants a long-running agent in an IDE, Replit is reasonable.

Card 6.5. Gnesis vs Cursor

What they are. The IDE that most serious developers in 2026 use. Heavy autocomplete, heavy in-editor agent. Their pricing (April 2026). Hobby free. Pro $20/mo ($16/mo annual). Pro+ $60/mo. Ultra $200/mo. Teams $40/user. Enterprise custom with SSO, audit, SCIM. What they do well. Best-in-class editor agent for existing codebases. Fast, precise, developer-loved. Where they fall short for our ICP.

Our winning play. "Cursor is the best way to edit software. Gnesis is the fastest way to bring software into existence. Our customers graduate from Gnesis into Cursor once they hire their first engineer." Frame Cursor as a complement, not a competitor. What we concede. A team with five experienced engineers and an existing codebase should be using Cursor, not Gnesis. Walk.

Card 6.6. Gnesis vs Claude Code

What they are. Anthropic's terminal-first coding agent. In April 2026 access was briefly restricted to Max ($100/mo minimum) before being partially reversed. Their pricing (April 2026, volatile). Bundled with Pro ($20/mo) under test, Max 5x ($100/mo), Team Premium ($100/seat, 5-seat min), Enterprise. Claude Code revenue reportedly approaching $2.5B annualised. What they do well. Deep terminal integration, great for senior engineers running multi-step agent loops. Strong at reasoning over existing repos. Where they fall short for our ICP.

Our winning play. "Claude Code is a hammer for developers. Gnesis is a factory for founders. We both love Anthropic; that is why Opus routes our architecture decisions." Turn the shared stack into a credibility point, not a conflict. What we concede. If the buyer is a senior engineer shipping a side project, Claude Code is great and we are the wrong tool.

Card 6.7. Gnesis vs Webflow

What they are. The dominant visual site builder, strongest for marketing sites and light CMS, with a mature enterprise motion. Their pricing (April 2026). Free starter. Business site plan from $39/month. Enterprise custom. What they do well. Marketing sites, CMS, designer-friendly visual control, mature enterprise compliance story. Where they fall short for our ICP.

Our winning play. "Webflow is the right answer for your marketing site. It is the wrong answer for your SaaS. Gnesis ships your product; Webflow dresses your homepage." Offer to integrate, not compete. What we concede. Marketing site, done.

Card 6.8. Gnesis vs Retool

What they are. Internal tools over existing databases. Dominant in the "ops dashboard" niche. Their pricing (April 2026). Free up to 5 standard + 5 end users. Team $10/standard user. Business $50/standard user. Enterprise custom. What they do well. Connect to your warehouse, build an internal dashboard, ship it to Ops. Fast for that niche. Where they fall short for our ICP.

Our winning play. "Retool is excellent for the tool your ops team uses every Tuesday. Gnesis is for the product your customers pay for every month. Different target user, different surface, different contract." Concede the internal-tool niche cleanly. What we concede. Internal tools on existing databases: Retool wins every time.

7. When we win (lead with these)

Mapped to the four deck ICPs and the four-phase GTM.

  1. Idea-rich, resource-poor founders (Phase 1, Q2 2026). Serial or second-time founders who want to avoid rebuilding auth, billing, tenancy, and compliance from scratch. Gnesis erases the first 90 days. ACV $500 to $2,000 per year.
  2. Corporate innovation and R&D (Phase 2, Q3 2026). Internal teams tired of explaining why an internal build is 18 months and still shaky. VPC-Vault, audit, 0 percent model markup at Enterprise. ACV $80K to $800K per year.
  3. Agencies and consultancies (Phase 3, Q4 2026). Shops that ship client work at volume. 20 percent white-label revenue share on base seat fee, standard quality floor on every build. ACV $30K to $60K plus rev share.
  4. Universities and bootcamps (Phase 4, Q1 2027). Cohort pricing at $50K to $150K per cohort. Outcome-reportable curriculum, real builds that ship.
  5. Lateral strength, not a fifth ICP: the six-month mark. Any of the four ICPs that has been burned by a demo that shipped and then broke. The line is: everyone is optimizing the demo; Gnesis optimizes the six-month mark.

8. When we lose (walk away honestly)

  1. Buyer wants a landing page. Send them to Webflow or Framer.
  2. Buyer is a senior engineer who just wants a better IDE agent. Cursor.
  3. Buyer needs a specific dashboard on an existing warehouse. Retool.
  4. Buyer wants a prototype for a pitch tomorrow and will throw it away next week. Bolt or Lovable.
  5. Buyer insists on full IDE control and wants to own every line. Gnesis is the wrong abstraction.

9. Objection handling

Ten objections you will hear, with responses.

9.1. "Bolt and Lovable are cheaper."

"They are. Their customers also rebuild auth, billing, tenancy, and compliance from scratch the moment they get a paying user. Our entry is $19 per seat plus AI at pass-through with a 20% markup; all-in a typical operator lands around $40-50 per seat per month. Our floor is multi-tenant, compliant, production. Different price because different product."

9.2. "I already have Cursor. Why do I need this?"

"Cursor edits code. Gnesis brings code into existence. Keep Cursor for when you hire your first engineer. Use Gnesis for the three months before that."

9.3. "Why voice? I prefer to type."

"Voice is default, not required. You can type to Shakeeb. The reason voice matters is latency: Shakeeb's voice round-trip target is 800ms p95, which is the point where conversation feels real. Typing to an agent still works; we just removed a friction layer."

9.4. "How do I trust an AI cofounder?"

"You do not. You trust the contract. Shakeeb operates under an explicit trust contract with approval gates on destructive actions: deploy, destroy, charge, provision, region-switch. You approve each one the first time, and configure standing approvals after that."

9.5. "What about data residency?"

"US is the default. EU and Middle East regions are on the Q2 2026 roadmap. Team+ customers choose region at provisioning. Enterprise can pin to VPC-Vault and bring their own cloud account."

9.6. "What about SOC 2?"

"SOC 2 Type I is targeted for Q3 2026. Type II is targeted for Q1 2027 and onward. We publish the timeline and the subprocessor list so nobody is guessing."

9.7. "Can I eject if I outgrow you?"

"Yes. The generated app is standard TypeScript, Next.js 16 + React 19, tRPC v11, Drizzle, per-tenant Postgres 16 via Coolify, BullMQ on Redis, Better Auth, Stripe, next-intl. Nothing proprietary in the runtime. If you leave, you take a repo plus a Docker compose plus the database dump plus the Coolify service descriptor. Any engineer can open and ship it."

9.8. "Who owns the code?"

"You do. Gnesis runs the factory; the output belongs to the customer. Aleph stays on our side. The generated product stays on yours."

9.9. "Why is Gnesis more expensive at Enterprise?"

"Enterprise is $50K/year with 0% markup on model and infra pass-through. At that tier you buy the contract, the isolation mode, the VPC-Vault, and the support. Starter, Team, and Business carry markup because we do the work of cost management on your behalf."

9.10. "What happens if a build fails?"

"We absorb the cost of any failure before the Test stage. After Test, failures pass through at cost with zero markup. This is a Gnesis-sourced rule from the pricing model, not a negotiation point."

10. Disqualification criteria (walk fast)

Walk if the buyer:

Walking away is a win. It keeps ICP sharp and keeps the success base clean.

11. Proof points (only what is verifiable)

Gnesis-sourced proof points, all from the source-of-truth spec folder:

Do not cite any number not in this list as a "Gnesis-sourced" fact.

12. Competitor sources (verified April 2026)

13. Open questions for v1

14. Change log

Gnesis · Gnesis Battlecard · private beta · 2026-04-23