RESPONSIBILITIES

- Design architecture for RAG systems, agentic workflows and LLM integrations
- Define standards for AI evaluation, model selection, prompt engineering and production readiness
- Review technical designs produced by AI Engineers before implementation
- Own the AI platform roadmap and communicate it to product and leadership teams
- Hands-on implementation on high-complexity features — not purely supervisory
- Mentor AI Engineers and Backend Engineers on AI-specific patterns
- Maintain awareness of model capabilities, API updates and relevant open-source developments

REQUIRED SKILLS

- Python — production-grade development and code review
- LLM API integration — OpenAI, Anthropic or Azure OpenAI. Streaming, function calling, structured outputs, error handling
- RAG architecture — has designed and shipped retrieval-augmented generation systems. Can articulate choices around chunking, embedding models, reranking and evaluation




- Agentic systems — has built multi-step agent pipelines with tool use, memory and error recovery
- LLM evaluation — automated test frameworks, hallucination detection, retrieval quality measurement
- Vector databases — Milvus, Pinecone, Weaviate, Chroma or pgvector in production
- FastAPI or equivalent — API design and review
- Docker, cloud (AWS/Azure/GCP), basic Kubernetes — can review and guide deployment decisions
- Data Systems – Hands-on with Databases, Data architecture and distributed data systems

PREFERRED BACKGROUND

- 10+ years of software engineering with at least 2 years on AI or ML systems
- Has led AI delivery teams, shipped at least one RAG system or agentic workflow in a production setting
- Experience in professional services, financial services or enterprise SaaS is an advantage

Fine-tuning experience (LoRA/QLoRA) is a plus, not a requirement

📌 AI Transformation Lead (Mumbai)
🏢 BDO India
📍 Mumbai