Founding AI Engineer

Founding AI Engineer

About the Role

We’re a stealth AI startup in Ahmedabad. As Founding AI Engineer, you will design and ship LLM-powered agents and RAG pipelines with rigorous evals, safety, and reliability, then shape the platform architecture across BE/FE and help hire the AI team.

Read this first: This is a startup role. In intense weeks, expect 10–12 hours/day. You’ll be a founding team member and will be rewarded for early contribution (performance bonuses). Onsite only (Ahmedabad).

What You’ll Do

• Design & implement LLM agents and RAG pipelines with measurable quality.
• Own MLOps: datasets, eval harnesses (quality/latency/cost), prompt governance.
• Productionize inference paths with guardrails, fallbacks, observability, and cost controls.
• Partner with Product on problem → spec → ship; run experiments and measure ROI.
• Influence system architecture across backend/frontend; set engineering standards.
• Create internal tooling for offline/online evals, red/blue tests, and A/B experiments.

What You’ll Bring

• 4–8+ years total engineering; 2+ years applied LLM/GPT work shipped to prod.
• Proven rigor with evals (precision/recall, factuality/hallucination rate), latency SLOs, and cost tracking.
• Strong in one primary stack: Python + PyTorch + FastAPI (or equivalent).
• Experience with retrieval, vector stores, prompt/tool orchestration, and observability.
• Bonus: retrieval tuning, tools/agents, finetuning/LoRA, RLHF; Web3/fintech exposure a plus.
• Onsite in Ahmedabad; comfortable with 10–12h launch weeks when needed.

Why Join Us

• Greenfield system—your decisions set the standard.
• Direct access to design partners; ship features used the same week.
• Competitive salary + performance bonuses; in-office, focused craft culture.

Tech / Stack

Python, PyTorch, FastAPI, OpenAI/Anthropic + local models, LangChain/LlamaIndex, Vector DB (pgvector/Milvus/FAISS), Postgres, Redis, Docker, Kubernetes, Observability (Grafana/Prometheus, OpenTelemetry), CI/CD, feature flags.

Must-Have Skills

LLMs · Agents · RAG · MLOps · Prompt Engineering · Evals · Python · PyTorch · FastAPI · Vector Databases · Postgres/Redis · Observability · Cost & Latency Optimization · Production AI

Nice-to-Have Skills

Finetuning/LoRA · RLHF · Feature Stores · Streaming/SSE · Experimentation Platforms · Security/PII handling · Web3/Fintech domain knowledge