Maya Chen
AI Engineer | LLM and RAG Systems | Full-Stack Development
Professional Summary
AI Engineer specializing in large language models, retrieval-augmented generation, Python, FastAPI, and TypeScript. Built production AI and full-stack systems serving 4,000+ users, improving workflow reliability and automating customer-facing operations. Experienced in shipping scalable backend services, evaluation workflows, and data-driven product features.
Technical Skills
Experience
- Architected a RAG market chat backend with Voyage AI embeddings, Pinecone vector search, and LLM reasoning for prediction products serving 4,000+ current users.
- Engineered Redis Pub/Sub and BullMQ indexing pipelines for 2 live data streams with automated Pinecone cleanup for resolved markets.
- Implemented an async draft-generation workflow with cached query hashing, Redis de-duplication, reroll feedback, and SSE streaming across 4 generation stages.
- Designed a title and thumbnail generation service with 3 safeguards: token-bucket rate limits, Redis response caching, and BullMQ queue processing.
- Developed a market intelligence agent combining Pinecone context and web search to produce summaries, sentiment signals, impact scores, and social drafts.
Projects
Tech Stack: Next.js, TypeScript, Groq, PostgreSQL, email automation, and sales data views.
- Shipped a shopping assistant for 4 customer workflows: product guidance, sentiment-aware pricing, instant quotes, and bulk-order email support.
Tech Stack: Python, FastAPI, Manim, Groq, Mistral, Docker, Cloudinary, Next.js, Render, and Vercel.
- Created a 5-step animation service that turns natural-language inputs into narrated videos through parsing, code generation, rendering, TTS narration, and cloud upload.
Tech Stack: Next.js, React, TypeScript, WebSockets, Redis, Google Maps, PostgreSQL, Prisma, and Docker.
- Launched a real-time ride dashboard with 3 live signals: demand density, geo updates, and fare-anomaly patterns to make regional pricing behavior visible.
Tech Stack: Next.js, TypeScript, Prisma, PostgreSQL, Turborepo, Docker, CI/CD, and Tailwind CSS.
- Delivered a full-stack wallet with 4 payment flows: add money, withdraw funds, transfer balance, and transaction history with webhook-driven bank automation.
Education
- Coursework: Artificial Intelligence, Machine Learning, Data Structures and Algorithms, Operating Systems, Computer Networks, and Object-Oriented Programming.
Certifications
Leadership and Open Source
- Merged pull requests across MERN projects covering validation, routing fixes, and reusable React state-management utilities.
- Coordinated a state-level 24-hour hackathon with 100+ participants, problem statements, mentoring, and cross-team communication.