# Project export: GENESIS AI This document was generated by HackStack to give an AI agent context about a hackathon project. Sections are labeled with their provenance; content marked as truncated was cut to keep this document small. ## Project metadata - Hackathon: OpenAI Build Week - Tagline: Genesis transforms a single mission into a complete AI-powered organization with structure, execution plans, worker assignments, validation, and deployment guidance in minutes. - Devpost: https://devpost.com/software/openai-03hjfs - GitHub: https://github.com/jagadiswarambati/GENESIS---ORGANISATION-AI - Video: https://www.youtube.com/embed/tuS3NOaaYP4?enablejsapi=1&hl=en_US&rel=0&start=&version=3&wmode=transparent - Team: 1 GitHub contributor(s) — jagadiswarambati (9 commits) ## Devpost submission (written by the team) ### Inspiration Organizations often spend weeks planning teams, workflows, documentation, and execution before real development begins. Existing AI tools generate responses, but they don't create complete operational organizations. Genesis was built to bridge that gap. ### What it does Genesis is an AI Organization Architect that transforms a single mission into a complete AI-powered operational organization. It automatically generates organizational structures, execution plans, worker assignments, project documentation, validation reports, deployment guidance, and exportable project foundations. ### How we built it Genesis was built during OpenAI Build Week using GPT-5.6 and Codex. The frontend is built with Next.js, React, TypeScript, and Tailwind CSS. The backend uses FastAPI with provider abstraction supporting OpenAI, Gemini, and Ollama. Codex accelerated architecture design, backend implementation, debugging, provider integration, documentation, and workflow optimization. ### Challenges we ran into One of the biggest challenges was coordinating multiple AI generation stages while keeping the interface responsive. We redesigned the workflow to support background AI execution, progressive updates, timeout handling, retries, and provider abstraction. ### Accomplishments we're proud of • Complete AI organization generation • Interactive Mission Control dashboard • Background AI execution • Multi-provider AI support • Automatic documentation generation • Exportable project foundation ### What we learned We learned how to combine structured AI reasoning with software architecture to build systems that generate organizations instead of isolated responses. We also learned how important workflow optimization and asynchronous execution are for user experience. ### What's next for Genesis Future work includes collaborative multi-user organizations, persistent organization memory, agent collaboration, cloud deployment, organization templates, and deeper enterprise integrations. ## README (from the GitHub repository) # Genesis > **Build Organizations. Not Prompts.** **Genesis** is an AI Organization Platform—an Organization OS for AI work—built for OpenAI Build Week. Give Genesis a mission and it designs a validated organization, turns its departments into an execution plan, coordinates specialist workers, captures shared knowledge, and assembles a reviewable project foundation. [![OpenAI Build Week](https://img.shields.io/badge/OpenAI-Build%20Week-412991?logo=openai&logoColor=white)](https://openai.com/) [![Next.js 15](https://img.shields.io/badge/Next.js-15-black?logo=next.js)](frontend/package.json) [![FastAPI](https://img.shields.io/badge/FastAPI-0.115+-009688?logo=fastapi&logoColor=white)](backend/pyproject.toml) [![Python 3.12+](https://img.shields.io/badge/Python-3.12+-3776AB?logo=python&logoColor=white)](backend/pyproject.toml) [![License: MIT](https://img.shields.io/badge/License-MIT-yellow.svg)](LICENSE) [> **Demo video:** ](https://youtu.be/tuS3NOaaYP4?si=ueCT_p442oZbnqJp) # 📸 Screenshots ## 🏠 Mission Input The entry point of Genesis where users describe their mission in natural language. ![Mission Input](docs/screenshots/home.png) --- ## ❓ Organization Questionnaire Genesis gathers additional context through a guided organization questionnaire before generating the final blueprint. ![Organization Questionnaire](docs/screenshots/questionaire.png) --- ## 🏗️ Organization Blueprint The generated organization is visualized as an interactive blueprint showing departments, relationships, and overall structure. ![Organization Blueprint](docs/screenshots/blueprint.png) --- ## 🧬 Organization DNA Genesis summarizes the organization's identity through its mission, culture, values, and strategic DNA. ![Organization DNA](docs/screenshots/dna.png) --- ## 🎯 Mission Control Mission Control provides a live operational view of organization generation, execution progress, workers, background AI tasks, and project readiness. ![Mission Control](docs/screenshots/mission-control.png) --- --- ## 📂 Project Structure Genesis automatically generates a structured project foundation that serves as the starting point for implementation. ![Project Structure](docs/screenshots/structure.png) --- ## 🛡️ Sandbox Verification Generated project foundations are safely verified before export to ensure structural integrity and implementation readiness. ![Sandbox Verification](docs/screenshots/sandbox.png) --- ## 📦 Project Export Export the generated organization, documentation, and project foundation for future development. ![Project Export](docs/screenshots/export.png) --- ## In 60 seconds Genesis is for work that is too broad for one prompt and too connected for a collection of isolated agents. A user begins with a mission such as “Build a secure AI-powered hospital operations platform.” Genesis creates a typed Organization Blueprint, sequences the newly created departments, generates deterministic tasks and worker assignments, manages task readiness, and enriches the operational view in the background. The result is not a chat transcript. It is a navigable organization with a mission brief, DNA, departments, workers, workflow state, memory, artifacts, a workspace, package, validation report, verification report, review, and deployment plan. ## Problem statement Complex initiatives do not fail because teams lack a single answer; they fail because research, strategy, design, engineering, quality, operations, and communication need to happen in the right order and share context. Planning that organization is slow, difficult to inspect, and usually trapped in documents and meetings. Most AI products begin with a prompt and return an answer. That interaction is useful, but it does not model responsibilities, dependencies, handoffs, institutional knowledge, or project readiness. Genesis treats the organization itself as the product surface: a mission becomes a structured operating model that can be inspected before work starts and followed while it progresses. ## Solution Genesis turns a plain-language mission into an operational organization without exposing a provider-specific workflow to the browser. The Organization Architect produces a Pydantic-validated blueprint; deterministic services make the resulting sequence and work graph reproducible; a provider-neutral execution layer can use Mock AI, OpenAI, Gemini, or local Ollama; and Mission Control keeps the organization useful while enrichment continues asynchronously. ```mermaid flowchart TD M[Mission input] --> A[Organization Architect] A --> B[Validated Organization Blueprint] B --> Q[Conversational organization questions] Q --> P[Blueprint and Organization Brief] P --> MC[Mission Control] MC --> EP[Execution Planner] EP --> TG[Task Generator] TG --> WA[Worker Assignment] WA --> WF[Workflow Engine] WF --> CE[Collaboration and ready-task execution] CE --> MEM[Organization Memory] CE --> ART[Project-foundation artifacts] ART --> WS[Project Workspace] WS --> PKG[ZIP package and manifest] PKG --> V[Validation and safe verification] PKG --> R[Project review and selective refinement] PKG --> D[Deployment plan] MEM --> DASH[Mission Dashboard] V --> DASH R --> DASH D --> DASH ``` ## Key features ### Organization Architect and Blueprint The Architect accepts a mission through `POST /api/v1/architect`, asks a selected structured-response provider for JSON, then validates the result against the `OrganizationBlueprint` schema before returning it. The blueprint contains the organization name and type, mission summary, culture, Organization DNA, departments, roles, execution strategy, capacity, risks, deliverables, and confidence score. Invalid provider output is never passed directly to the UI. The Mission Initiation experience deliberately keeps the mission central. It moves from a focused mission input through a preparation animation and a one-question-at-a-time organization conversation to a React Flow blueprint, DNA, and an approval brief. Mission Control is reached only after the user approves that summary. ### Deterministic planning, tasks, and staffing The Execution Planner orders the departments in a validated blueprint using department and mandate keywords—research, strategy, design, engineering, quality, operations, then marketing—while retaining unmatched departments. The Task Generator expands each phase using deterministic department templates. Worker Assignment creates role-specific workers and attaches tasks using department and task-type rules. This separation matters: a provider may change, but the planned organization, task graph, and worker assignment remain explainable and repeatable. It also avoids asking a model to regenerate information Genesis already has. ### Workflow Engine and provider-neutral execution Every task has a typed lifecycle: `Pending`, `Ready`, `Running`, `Completed`, `Failed`, or `Blocked`. Only work without unsatisfied dependencies is initially ready; the workflow service refreshes dependent work when prerequisites complete. The Execution Engine consumes only ready workflow tasks. It depends on the `AIProvider` interface rather than a vendor SDK, so the same workflow can use the deterministic Mock provider, OpenAI, Gemini, or Ollama. Provider selection is one environment value, not a frontend or API contract change. ### Collaboration and organization memory The Collaboration Engine records phase-scoped messages with a sender, optional receiver, message type, related task, and timestamp. Execution context includes organization memory, relevant artifacts, prior decisions, and task-related conversation history. Completed work produces typed memory entries; artifact, workspace, and refinement references are also retained. Mission Control exposes this as Team Collaboration and Organization Memory instead of hiding institutional context inside individual model cal [README truncated for size] ## Detected evidence (automated analysis) Indexed codebase: 175 recognized source files, 672 KB. - CSS (language) — detected in the code - FastAPI (technology) — detected in the code - Next.js (technology) — detected in the code - OpenAI (technology) — detected in the code - Python (language) — detected in the code - React (technology) — detected in the code - Tailwind CSS (technology) — detected in the code - TypeScript (language) — detected in the code - Docker (technology) — claimed on Devpost, not found in the code - Google Gemini (technology) — claimed on Devpost, not found in the code - Ollama (technology) — claimed on Devpost, not found in the code ## Codebase structure (from repository index) ### Files (120 of 198) ``` .editorconfig .env.example .github/workflows/ci.yml .gitignore .prettierignore .prettierrc assets/README.md backend/.dockerignore backend/app/__init__.py backend/app/api/__init__.py backend/app/api/v1/architect.py backend/app/api/v1/artifacts.py backend/app/api/v1/collaboration.py backend/app/api/v1/deployment.py backend/app/api/v1/execution_planner.py backend/app/api/v1/execution.py backend/app/api/v1/packaging.py backend/app/api/v1/project_review.py backend/app/api/v1/router.py backend/app/api/v1/system_health.py backend/app/api/v1/task_generator.py backend/app/api/v1/validation.py backend/app/api/v1/verification.py backend/app/api/v1/worker_assignment.py backend/app/api/v1/workflow.py backend/app/api/v1/workspace.py backend/app/core/__init__.py backend/app/core/config.py backend/app/core/database.py backend/app/core/errors.py backend/app/main.py backend/app/models/__init__.py backend/app/repositories/__init__.py backend/app/schemas/__init__.py backend/app/schemas/architect.py backend/app/schemas/artifact.py backend/app/schemas/base.py backend/app/schemas/collaboration.py backend/app/schemas/deployment.py backend/app/schemas/errors.py backend/app/schemas/execution_plan.py backend/app/schemas/execution.py backend/app/schemas/memory.py backend/app/schemas/packaging.py backend/app/schemas/review.py backend/app/schemas/system_health.py backend/app/schemas/task_generator.py backend/app/schemas/validation.py backend/app/schemas/verification.py backend/app/schemas/worker_assignment.py backend/app/schemas/workflow.py backend/app/schemas/workspace.py backend/app/services/__init__.py backend/app/services/architect_parser.py backend/app/services/architect_prompt.py backend/app/services/artifact_catalog.py backend/app/services/artifact_generation.py backend/app/services/collaboration_engine.py backend/app/services/collaborative_execution.py backend/app/services/deployment_generation.py backend/app/services/execution_engine.py backend/app/services/execution_planner.py backend/app/services/gemini_client.py backend/app/services/ollama_client.py backend/app/services/openai_client.py backend/app/services/organization_architect.py backend/app/services/organization_memory.py backend/app/services/packaging_engine.py backend/app/services/project_review.py backend/app/services/providers/__init__.py backend/app/services/providers/base.py backend/app/services/providers/collaborative.py backend/app/services/providers/factory.py backend/app/services/providers/gemini.py backend/app/services/providers/mock.py backend/app/services/providers/ollama.py backend/app/services/providers/openai.py backend/app/services/review_providers/__init__.py backend/app/services/review_providers/base.py backend/app/services/review_providers/factory.py backend/app/services/review_providers/gemini.py backend/app/services/review_providers/mock.py backend/app/services/review_providers/ollama.py backend/app/services/review_providers/openai.py backend/app/services/review_providers/schemas.py backend/app/services/system_health.py backend/app/services/task_generator.py backend/app/services/validation_engine.py backend/app/services/verification_engine.py backend/app/services/worker_assignment.py backend/app/services/workflow_engine.py backend/app/services/workspace_engine.py backend/app/utilities/__init__.py backend/Dockerfile backend/pyproject.toml backend/pytest.ini backend/README.md backend/requirements.txt backend/tests/conftest.py backend/tests/test_collaborative_execution.py backend/tests/test_release_validation.py backend/tests/test_workspace_quality.py backend/uv.lock database/alembic.ini database/alembic/env.py database/alembic/script.py.mako database/alembic/versions/README.md database/README.md docker-compose.yml docs/architecture.md docs/design-system.md docs/organization-architect.md docs/release-validation-report.md docs/roadmap.md docs/tech-stack.md docs/vision.md eslint.config.mjs frontend/.dockerignore frontend/app/globals.css frontend/app/launch/page.tsx [78 more files omitted for size] ``` ### Dependencies - backend/pyproject.toml: alembic@>=1.14,<2.0, asyncpg@>=0.30,<1.0, fastapi@>=0.115,<1.0, google-genai@>=2.0,<3.0, httpx@>=0.28,<1.0, openai@>=1.0,<2.0, pgvector@>=0.3,<1.0, pydantic-settings@>=2.7,<3.0, sqlalchemy[asyncio]@>=2.0,<3.0, uvicorn[standard]@>=0.32,<1.0 - backend/requirements.txt: alembic@>=1.14,<2.0, asyncpg@>=0.30,<1.0, fastapi@>=0.115,<1.0, google-genai@>=2.0,<3.0, httpx@>=0.28,<1.0, openai@>=1.0,<2.0, pgvector@>=0.3,<1.0, pydantic-settings@>=2.7,<3.0, sqlalchemy[asyncio]@>=2.0,<3.0, uvicorn[standard]@>=0.32,<1.0 - frontend/package.json: @types/node@^22.10.1, @types/react@^19.0.2, @types/react-dom@^19.0.2, @xyflow/react@^12.8.0, autoprefixer@^10.4.20, class-variance-authority@^0.7.1, clsx@^2.1.1, eslint@^9.17.0, eslint-config-next@^15.5.0, framer-motion@^12.9.2, lucide-react@^0.468.0, next@^15.5.0, postcss@^8.4.49, react@^19.0.0, react-dom@^19.0.0, tailwind-merge@^2.5.5, tailwindcss@^3.4.17, typescript@^5.7.2 - package.json: @eslint/eslintrc@^3.3.1, prettier@^3.5.3, prettier-plugin-tailwindcss@^0.6.11 ### Recent commits (newest first) - Replace demo video placeholder with actual link - Update README.md - added screenshots - Enhance README with feature descriptions and screenshots - Update GENESIS_OLLAMA_MODEL to include alternative option - final readme - version 2 - second commit - Initial commit ## Key source files (fetched from GitHub, selected and truncated for size) ### docs/roadmap.md ```markdown # Roadmap ## Milestone 1 — Project Initialization Establish the monorepo, frontend and backend foundations, database migration structure, local development environment, and project documentation. ## Future milestones - Organization modeling and versioned structure - Organization Architect and organization setup - Mission planning and governed execution - Institutional memory and artifact lineage - Organization health, learning, and evolution insights The order and scope of future milestones remain product decisions. This document records the agreed strategic direction without pre-committing implementation. ``` ### docs/vision.md ```markdown # Genesis Vision ## Build Organizations. Not Prompts. Genesis is an AI Organization Platform and the Organization OS for AI Work. It gives people a way to design, operate, learn from, and evolve AI-powered organizations around meaningful missions. The long-term product is built around four promises: 1. **Design** organizations with purposeful structure and behavior. 2. **Operate** missions through coordinated, governed work. 3. **Learn** from outcomes through institutional knowledge. 4. **Evolve** organizations using measured performance. Genesis is not a collection of isolated prompts or agents. It is a system for making organizational intelligence visible, governable, and compounding. ``` ### requirements.txt ``` # Compatibility entry point. Genesis uses uv and backend/pyproject.toml as the # authoritative Python dependency manifest. -r backend/requirements.txt ``` ### package.json ``` { "name": "genesis", "version": "0.1.0", "private": true, "description": "Genesis — Build Organizations. Not Prompts.", "workspaces": [ "frontend" ], "scripts": { "dev": "npm run dev --workspace=@genesis/frontend", "build": "npm run build --workspace=@genesis/frontend", "lint": "npm run lint --workspace=@genesis/frontend", "typecheck": "npm run typecheck --workspace=@genesis/frontend", "format": "prettier --write .", "format:check": "prettier --check ." }, "devDependencies": { "@eslint/eslintrc": "^3.3.1", "prettier": "^3.5.3", "prettier-plugin-tailwindcss": "^0.6.11" }, "packageManager": "npm@10.9.2" } ``` ### docker-compose.yml ```yaml services: frontend: build: context: ./frontend ports: - "3000:3000" volumes: - ./frontend:/app - frontend_node_modules:/app/node_modules environment: WATCHPACK_POLLING: "true" NEXT_PUBLIC_API_BASE_URL: ${NEXT_PUBLIC_API_BASE_URL:-http://localhost:8000} depends_on: backend: condition: service_started backend: build: context: ./backend ports: - "8000:8000" volumes: - ./backend:/app environment: GENESIS_ENVIRONMENT: ${GENESIS_ENVIRONMENT:-development} GENESIS_DATABASE_URL: ${GENESIS_DATABASE_URL:-postgresql+asyncpg://genesis:genesis@postgres:5432/genesis} GENESIS_REDIS_URL: ${GENESIS_REDIS_URL:-redis://redis:6379/0} GENESIS_FRONTEND_ORIGIN: ${GENESIS_FRONTEND_ORIGIN:-http://localhost:3000} GENESIS_AI_PROVIDER: ${GENESIS_AI_PROVIDER:-mock} GENESIS_OPENAI_MODEL: ${GENESIS_OPENAI_MODEL:-gpt-5.6} OPENAI_API_KEY: ${OPENAI_API_KEY:-} GENESIS_GEMINI_MODEL: ${GENESIS_GEMINI_MODEL:-gemini-3.5-flash} GEMINI_API_KEY: ${GEMINI_API_KEY:-} GENESIS_OLLAMA_BASE_URL: ${GENESIS_OLLAMA_BASE_URL:-http://host.docker.internal:11434} GENESIS_OLLAMA_MODEL: ${GENESIS_OLLAMA_MODEL:-llama3.2:3b} depends_on: postgres: condition: service_healthy redis: condition: service_started postgres: image: pgvector/pgvector:pg16 ports: - "5432:5432" environment: POSTGRES_DB: ${POSTGRES_DB:-genesis} POSTGRES_USER: ${POSTGRES_USER:-genesis} POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:-genesis} volumes: - postgres_data:/var/lib/postgresql/data healthcheck: test: ["CMD-SHELL", "pg_isready -U $$POSTGRES_USER -d $$POSTGRES_DB"] interval: 5s timeout: 5s retries: 10 redis: image: redis:7-alpine ports: - "6379:6379" volumes: - redis_data:/data volumes: frontend_node_modules: postgres_data: redis_data: ``` ### frontend/Dockerfile ``` FROM node:22-alpine WORKDIR /app COPY package.json ./ RUN npm install COPY . . EXPOSE 3000 CMD ["npm", "run", "dev", "--", "-H", "0.0.0.0"] ``` ### backend/requirements.txt ``` # Compatibility export for environments that install with pip instead of uv. alembic>=1.14,<2.0 asyncpg>=0.30,<1.0 fastapi>=0.115,<1.0 google-genai>=2.0,<3.0 httpx>=0.28,<1.0 openai>=1.0,<2.0 pgvector>=0.3,<1.0 pydantic-settings>=2.7,<3.0 sqlalchemy[asyncio]>=2.0,<3.0 uvicorn[standard]>=0.32,<1.0 ``` ### backend/Dockerfile ``` FROM python:3.12-slim ENV PYTHONDONTWRITEBYTECODE=1 \ PYTHONUNBUFFERED=1 \ UV_PROJECT_ENVIRONMENT=/opt/venv WORKDIR /app RUN pip install --no-cache-dir uv COPY pyproject.toml README.md ./ RUN uv sync --no-dev COPY app ./app EXPOSE 8000 CMD ["/opt/venv/bin/uvicorn", "app.main:app", "--host", "0.0.0.0", "--port", "8000", "--reload"] ``` ### backend/pyproject.toml ``` [project] name = "genesis-backend" version = "0.1.0" description = "Backend foundation for Genesis, the Organization OS for AI Work." readme = "README.md" requires-python = ">=3.12" dependencies = [ "alembic>=1.14,<2.0", "asyncpg>=0.30,<1.0", "fastapi>=0.115,<1.0", "google-genai>=2.0,<3.0", "httpx>=0.28,<1.0", "openai>=1.0,<2.0", "pgvector>=0.3,<1.0", "pydantic-settings>=2.7,<3.0", "sqlalchemy[asyncio]>=2.0,<3.0", "uvicorn[standard]>=0.32,<1.0", ] [dependency-groups] dev = [ "pytest>=8.3,<9.0", "ruff>=0.8,<1.0", ] [tool.uv] package = false [tool.ruff] target-version = "py312" line-length = 100 [tool.ruff.lint] select = ["E", "F", "I", "UP", "B"] [tool.ruff.format] quote-style = "double" indent-style = "space" ``` ### frontend/package.json ``` { "name": "@genesis/frontend", "version": "0.1.0", "private": true, "scripts": { "dev": "next dev", "build": "next build", "start": "next start", "lint": "eslint .", "typecheck": "tsc --noEmit" }, "dependencies": { "@xyflow/react": "^12.8.0", "class-variance-authority": "^0.7.1", "clsx": "^2.1.1", "framer-motion": "^12.9.2", "lucide-react": "^0.468.0", "next": "^15.5.0", "react": "^19.0.0", "react-dom": "^19.0.0", "tailwind-merge": "^2.5.5" }, "devDependencies": { "@types/node": "^22.10.1", "@types/react": "^19.0.2", "@types/react-dom": "^19.0.2", "autoprefixer": "^10.4.20", "eslint": "^9.17.0", "eslint-config-next": "^15.5.0", "postcss": "^8.4.49", "tailwindcss": "^3.4.17", "typescript": "^5.7.2" } } ``` [165 more indexed source files omitted to keep this export small. The full file list is in the Codebase structure section above.]