Darwin

Darwin creates a Digital Twin of a founder and uses an AI executive board to stress-test startup ideas against real constraints, then generates execution-ready startup plans.

Comment

Inspiration

Every year, thousands of founders spend months building startups that were never the right fit for them. Most startup tools focus entirely on the idea while ignoring the person building it.

A great startup for one founder can be a terrible startup for another founder due to differences in skills, capital, network, risk tolerance, execution speed, and experience.

We wanted to answer a different question:

"What is the best startup for this specific founder?"

That idea led to Darwin.

Darwin creates a Digital Twin of a founder and uses an AI Executive Board to evaluate startup ideas against the founder's actual strengths and constraints before valuable time, money, and effort are wasted.

What it does

Darwin is an AI-powered Founder Intelligence Platform.

The platform begins by creating a Digital Twin of the founder through a structured onboarding process. Darwin analyzes the founder's technical skills, execution capability, budget, network, risk profile, goals, strengths, weaknesses, and constraints.

Once the Digital Twin is created, founders can pitch startup ideas to an AI Executive Board composed of:

  • CEO Agent
  • CFO Agent
  • CTO Agent
  • CMO Agent
  • CPO Agent

The board conducts a structured three-round debate:

  1. Initial Position Analysis
  2. Cross Examination
  3. Final Vote

The board can decide to:

  • Proceed
  • Pivot
  • Reject

After approval, Darwin automatically generates:

  • Founder Intelligence Report
  • Opportunity Analysis
  • Startup Blueprint
  • MVP Scope
  • Financial Roadmap
  • Execution Plan

Using GitLab MCP integration, Darwin then creates a GitLab repository, milestones, epics, issues, and execution workflows that transform strategy into actionable development tasks.

How we built it

Darwin was built using a modern AI-first architecture.

Frontend

  • Next.js
  • TypeScript
  • Tailwind CSS
  • Framer Motion

Backend

  • FastAPI
  • Python
  • Pydantic

AI Layer

  • Google Gemini
  • Multi-Agent Executive Board System
  • Digital Twin Engine

Google Cloud Services

  • Gemini
  • Vertex AI
  • Firebase Authentication
  • Firestore
  • Cloud Run
  • Cloud Storage

Database

  • MongoDB
  • Firestore

Partner Integration

  • GitLab MCP

The platform uses Gemini to synthesize founder profiles into Digital Twins and orchestrate reasoning across multiple executive agents. GitLab MCP enables Darwin to convert strategic decisions directly into development workflows and repository structures.

Challenges we ran into

One of the biggest challenges was preventing the AI system from simply agreeing with the founder.

Most AI assistants naturally try to be helpful and supportive. Darwin needed to be critical, analytical, and willing to challenge assumptions.

Another challenge was designing a consistent multi-agent debate framework where each executive maintained a unique perspective while contributing toward a final board decision.

We also had to translate high-level strategic recommendations into practical execution workflows through GitLab, ensuring outputs were not only intelligent but actionable.

Accomplishments that we're proud of

We successfully built a complete founder-to-execution workflow.

Key accomplishments include:

  • Creating a Founder Digital Twin system
  • Building a multi-agent executive board architecture
  • Implementing structured AI debates
  • Generating founder-specific startup recommendations
  • Producing financial and execution roadmaps
  • Integrating GitLab MCP for automated project planning
  • Transforming strategy directly into development workflows

Most importantly, we built a system that helps founders make better decisions instead of simply generating more ideas.

What we learned

Throughout this project, we learned that founder-market fit is often more important than idea quality.

We discovered that effective AI agents require strong constraints, specialized roles, and structured reasoning rather than relying on a single powerful model.

We also learned that founders value honest criticism more than validation when making high-stakes decisions.

Finally, we gained valuable experience building systems that combine Digital Twins, multi-agent orchestration, execution planning, and developer workflows into a single product experience.

What's next for Darwin

Our vision is to evolve Darwin into a complete Founder Operating System.

Future versions will:

  • Continuously update the Digital Twin using real-world execution data
  • Integrate GitHub, GitLab, revenue, and product analytics
  • Add investor and mentor agents to the boardroom
  • Support team-level Digital Twins for startups
  • Track founder progress over time
  • Generate increasingly personalized strategic recommendations

Long term, Darwin aims to become the intelligence layer that helps founders make better decisions throughout the entire lifecycle of building a company.

Because the best startup is not always the one with the best idea.

It's the one the founder is most likely to successfully build.

Built With

Share this project:

Updates