Project Info

Genesis

Devpost

Inspiration

I've always loved startup ideas, but I realized that turning an idea into a real product takes much more than just coding. You need research, product planning, technical design, deployment, and constant iteration. Most AI tools only help with one piece of that journey. So I built Genesis - an autonomous AI organization that works like a real startup team, helping founders go from a single idea to a deployed SaaS product.

What it does

Genesis transforms a one-line startup idea into a complete SaaS venture. Instead of relying on a single AI assistant, it uses 18 specialized AI agents that collaborate to critique the idea, research the market, define the product, design the architecture, generate an MVP, deploy it to GitHub, Render, and Vercel, and continuously improve it through feedback. Every stage produces structured, traceable outputs while keeping humans in control of important decisions.

How we built it

Genesis is built with Next.js, FastAPI, LangGraph, and SQLite. The frontend streams every AI agent's progress in real time, while the backend orchestrates the workflow, stores project memory, maintains a knowledge graph, and integrates AI providers like Groq, OpenRouter, and optional Tavily for live web research. I also built automated deployment so Genesis can generate a GitHub repository, deploy the backend to Render, the frontend to Vercel, and even create an investor pitch.

Challenges we ran into

The biggest challenge was making multiple AI agents work together instead of behaving like separate chatbots. I also spent a lot of time handling API rate limits, provider failures, live progress streaming, deployment automation, and making sure every decision remained transparent and traceable.

Accomplishments we're proud of

The thing I'm most proud of is that Genesis feels like a real AI startup team rather than another AI chat interface. Seeing an idea go from evaluation to research, architecture, deployment, and finally a live application—all inside one workflow—was incredibly rewarding.

What we learned

This project taught me that building AI applications isn't just about using powerful models. It's about designing reliable systems where multiple AI agents can collaborate, share context, and produce meaningful outcomes while keeping humans in the loop.

What's next

I'm planning to make Genesis even more autonomous by improving code generation, expanding the knowledge graph, strengthening long-term memory, and adding more specialized AI roles. The long-term vision is to make Genesis an AI co-founder that can help anyone turn an idea into a real software business.

Analysis

Compare with all teams

View

Metric

Figures cover GitHub contributors during the hackathon window. A co-authored commit counts in full for each author, so per-member totals add up to more than the whole-team figures.

Technology

Found in codeClaimed only
  • CSSIn code
  • FastAPIIn code
  • HTMLIn code
  • LangChainIn code
  • Next.jsIn code
  • OpenAIIn code
  • PythonIn code
  • ReactIn code
  • Tailwind CSSIn code
  • TypeScriptIn code
  • Node.jsClaimed

10 of 11 appear in the indexed code. 1 claimed on Devpost could not be matched to code, which may simply mean the tool leaves no trace in the repository.

AI coding agents

  • Claude CodeConfig
  • CodexConfig

Detected from committed agent config files and commit authorship. Absence of a signal is not proof an agent was unused.

Codebase size

Source size

941 KB

Source files

117

Counts recognized source files only; vendored directories, binaries and lockfiles are excluded, so this is smaller than the repository on disk.

0 stars