Marque

AI-powered web app for personalized learning and document Q&A.

Marque turns the documents you already have — lecture notes, research papers, manuals, meeting transcripts — into an interactive learning surface. Upload a document, ask a question in natural language, and Marque returns a grounded answer with citations, follow-up suggestions, and lightweight study aids generated on the fly.

Built by a two-person team starting August 2024.

---

What's inside

• AI Q&A engine — Retrieval-Augmented Generation pipeline over a Pinecone vector store, with OpenAI embeddings and a Python evaluation harness that benchmarks answer quality on every release.
• Web app — Next.js (App Router) + TypeScript frontend on a custom Tailwind design system, with Framer Motion for animation and WCAG-compliant accessibility primitives.
• Backend — FastAPI + asyncio, PostgreSQL for application data, S3-backed media, deployed on AWS ECS Fargate behind CloudFront.
• Operations — GitHub Actions CI/CD with staged deploys, Docker containerization, Datadog metrics, Sentry error tracking.

Architecture

                  ┌─────────────────────────┐
                  │   Next.js (Vercel/CF)   │
                  │   App Router + Tailwind │
                  └────────────┬────────────┘
                               │ HTTPS
                               ▼
                  ┌─────────────────────────┐
                  │   FastAPI (ECS Fargate) │
                  │   asyncio + uvicorn     │
                  └──┬──────────────────┬───┘
                     │                  │
              ┌──────▼──────┐    ┌──────▼──────┐
              │  Postgres   │    │  Pinecone   │
              │  (RDS)      │    │  (vectors)  │
              └─────────────┘    └──────┬──────┘
                                        │
                                 ┌──────▼──────┐
                                 │  OpenAI API │
                                 │  (LLM + emb)│
                                 └─────────────┘

Repo layout

marque-web/
├── backend/        FastAPI service — Q&A, document ingestion, eval
├── frontend/       Next.js web app — landing + Q&A interface
├── .github/        CI/CD workflows
├── docker-compose.yml    local dev orchestration
└── README.md       you are here

Local development

Requires Docker, Node 20+, and Python 3.11+.

# Start backend + postgres
docker compose up -d postgres
cd backend && pip install -r requirements.txt && uvicorn app.main:app --reload

# Frontend (separate terminal)
cd frontend && npm install && npm run dev

Open http://localhost:3000.

Environment

Copy .env.example to .env and fill in:

• OPENAI_API_KEY — embedding + chat completion
• PINECONE_API_KEY / PINECONE_INDEX — vector store
• DATABASE_URL — Postgres connection string
• SENTRY_DSN — error tracking (optional in dev)

Status

Actively developed. Roadmap and open issues live in the issues tab.

License

MIT — see LICENSE.