🔗 Live Demo  ·  GitHub Repo

_{Hosted on Render & Vercel — backend sleeps after inactivity; initial cold-start may take 30–45s.}

What is HACKORA?

HACKORA is an AI-powered hackathon companion designed to take builders from early problem statements to judge-defended projects with zero guesswork.

"From Ideas to Innovation."
Your AI-powered workspace to research, plan, build, and defend hackathon projects.

Unlike superficial chatbots or wrapper tools, HACKORA is structured into two purpose-built execution stages:

---

🚀 Stage 1 — Explore & Plan

Transform raw problem statements or early innovation ideas into structured, build-ready project foundations:

• 📋 Forensic Problem Statement Breakdown: Ingest PDFs, DOCX, TXT, or markdown briefs. HACKORA extracts requirements, constraints, target users, and success criteria without hallucinating unmentioned facts.
• 💡 Open Innovation Discovery: Research competitors and existing approaches, discover unaddressed user pain points, and surface defensible Unique Selling Propositions (USPs).
• 💬 Conversational AI Workspace (BYOK): Multi-LLM provider support (OpenAI, Anthropic Claude, Google Gemini, Moonshot Kimi, Groq, OpenRouter). Bring your own API key with secure client-side encryption and zero backend key logging.
• 📐 On-Demand Technical Blueprints: Generate 3-way solution comparison matrices, Mermaid architecture diagrams, database schemas, API specs, and full pitch deck blueprints with golden demo paths.

---

🛡️ Stage 2 — Analyze & Defend

Paste any GitHub repository URL to verify what your project actually implements and prepare for technical judging:

• 🗂️ Navigable File Structure: Instant file tree filtering out lockfiles and build noise.
• 🤖 AI File Explanations: Cached plain-English module explanations.
• 🔍 Hybrid Semantic Search (Ask HACKORA): Keyword (full-text BM25) + dense vector embeddings fused via Reciprocal Rank Fusion (RRF) for grounded code answers citing files and lines.
• 🗺️ Interactive Architecture Map: Live React Flow dependency graph highlighting critical paths and churn hotspots.
• 📊 Code Quality Score & Blast Radius: Churn, complexity, and transitive dependency analysis ("what breaks if I change this?").
• ⚖️ Forensic Requirement Audit & Defense Prep: Verify whether features in the problem statement exist in code and simulate rigorous judge cross-examinations.

---

🏗️ Architecture

                    ┌─────────────────────────┐
                    │    HACKORA Workspace    │
                    └────────────┬─────────────┘
                                 ▼
         ┌───────────────────────┴────────────────────────┐
         ▼                                                ▼
┌─────────────────────────┐                      ┌─────────────────────────┐
│   Stage 1: Explore      │                      │   Stage 2: Defend       │
│   & Plan Workspace      │                      │   & Analyze Codebase    │
└────────────┬────────────┘                      └────────────┬────────────┘
             ▼                                                ▼
┌─────────────────────────┐                      ┌─────────────────────────┐
│ Multi-LLM BYOK Gateway  │                      │ AST Parser & Git Mining │
│ (OpenAI, Claude, Groq,  │                      │ (tree-sitter, GitPython,│
│  Gemini, Kimi, Router)  │                      │  sentence-transformers) │
└────────────┬────────────┘                      └────────────┬────────────┘
             ▼                                                ▼
┌─────────────────────────┐                      ┌─────────────────────────┐
│ Structured Blueprints,  │                      │ Supabase pgvector + RRF │
│ Diagrams, Pitch Decks   │                      │ Grounded Q&A + Audit    │
└─────────────────────────┘                      └─────────────────────────┘

⚙️ Tech Stack

Layer	Technology
Frontend	React 19 · Vite · TypeScript · Tailwind CSS · React Flow · Lucide
Backend	Python 3.12 · FastAPI · Uvicorn
Database	Supabase (PostgreSQL + pgvector + GIN Full-Text Indexing)
AST Parsing	tree-sitter (Python, JS, TS/TSX)
Embeddings	sentence-transformers (all-MiniLM-L6-v2)
AI Providers	OpenAI · Anthropic · Google Gemini · Moonshot AI · Groq · OpenRouter
Git Analysis	GitPython · AST import resolution

🚀 Getting Started

# Clone the repository
git clone https://github.com/<your-username>/DevLens-AI.git
cd DevLens-AI

# Backend setup
cd backend
python -m venv .venv
.venv\Scripts\Activate.ps1   # Windows (.venv/bin/activate on macOS/Linux)
pip install -r requirements.txt
cp .env.example .env          # Add your Supabase + Groq/AI provider keys
python app/main.py

# Frontend setup (separate terminal)
cd frontend
npm install
cp .env.example .env          # Set VITE_API_BASE_URL=http://localhost:8000
npm run dev

☁️ Deployment

• Frontend: Vercel
• Backend: Render

HACKORA — From Ideas to Innovation.