AI/LLM Intern – LangChain, LangGraph, Firebase

Working experience: No Experience We're looking for brilliant, obsessed interns (or professionals) who want to build the future of personalized AI. You should be willing to learn anything, solve real problems, and ship work that touches real users. This is not a corporate internship. This is startup life. You'll be coding, researching, building, testing, failing, and pushing boundaries—every day. What You'll Do Build and test agents using LangChain, LangGraph, and LLM APIs (Gemini, OpenAI, Claude, etc.) Help improve AI behavior through prompt engineering, fine-tuning, and feedback loops Work with multi-agent systems that simulate personality, emotion, and tone Create visualizations, charts, or logs that help us debug conversations Contribute to integrations with Firebase, Google Cloud Functions, and Vector DBs Help prototype skincare tools using Python, JS, and ML tooling What You Must Know (or Learn Fast) Frameworks & Tools: Python (LangChain, FastAPI, Pydantic, async/await logic) LangChain / LangGraph Google Cloud Functions (Node.js / Python) Firebase (Firestore, Realtime DB, Cloud Functions) Pinecone or similar vector DBs (We use multilingual-e5-large) JavaScript (ES6+) – for minor app/backend integrations Git/GitHub – commit history should reflect real thinking Core Logic & Concepts: How to model multi-agent AI systems with role-based routing Building and using RAG pipelines (retrieval-augmented generation) Embedding search and similarity matching Session management and memory (short-term vs long-term memory in LLM apps) Structured logging, user feedback handling, and prompt-response tracking Human-in-the-loop design and agent error recovery What We're Looking For Obsession with learning, hacking, and pushing LLMs to their limit No ego. Willing to do data cleaning, debugging, and brainstorms Clear communication — can explain your logic, even under pressure Passion for AI, or just helping real humans with software Bonus Points Experience with Face/GAN/3D ML pipelines Past experience building personal GPTs, agents, or automations Having shipped anything real that someone uses (your own tools count)