🚀 Remote GenAI / LLM Engineer in training — building LLM Apps, RAG Systems & AI Agents
🎯 Goal: Remote GenAI / LLM Engineer job (RAG, agents) in the next 3–6 months
🧠 Currently: LLM Post-Training Intern (RLHF) @ Ethara AI
🔥 Focus: RLHF, Prompt Engineering, RAG, LangChain, Groq API
🌱 Building: Production-style AI applications with Python, FastAPI, Streamlit, React & Node.js
I'm Sufiyan Khan, a remote GenAI / LLM engineer in training focused on practical LLM-powered applications, RAG systems, and AI automation tools.
At Ethara AI, I work as an LLM Post-Training Intern (RLHF) where I:
- Evaluate and justify 50+ LLM prompt–response pairs daily
- Judge answers on quality, reasoning and alignment
- Provide feedback used in RLHF / fine-tuning datasets to improve model behavior
So I’m learning AI from both sides:
- Training side (RLHF): how models behave, fail, and get aligned
- Application side: how to turn LLMs into real products (RAG apps, agents, APIs)
I’m ready to learn whatever it takes (backend, system design, DevOps, etc.) to ship reliable GenAI systems end‑to‑end.
- Daily hands-on experience with RLHF workflows, not just calling LLM APIs
- Understand LLM behavior from evaluation + alignment, not only from tutorials
- Actively building deployable GenAI products (RAG, support bots, tools)
- Treating this like a serious career: clear roadmap, daily learning, and public projects
Languages: Python, JavaScript
Frontend: Streamlit, React
Backend / APIs: FastAPI, Node.js, REST APIs
LLM / GenAI: Groq API, Prompt Engineering, RLHF, (learning) LangChain
RAG / Retrieval: ChromaDB, embeddings, vector search (in progress)
Tools & Dev: Git, GitHub, VS Code, dotenv, Hugging Face Spaces
- RLHF & LLM Evaluation: reward models, preference data, safety and alignment basics
- LLM Systems: RAG architectures, tools/agents, evaluation, safety
- System Design for GenAI Apps: architecture diagrams, caching, logging/monitoring, cost & latency trade-offs
- Full-Stack for GenAI: using React & Node/FastAPI to ship real-world AI products
- Classic ML (selectively): regression, classification, metrics – using Google ML Crash Course and scikit-learn, only as needed for understanding LLM pipelines