Gen AI Engineer – LLM & RAG (Banking)

About the role

We are looking for a Gen AI Engineer to design and deliver retrieval‑augmented generation (RAG) solutions for banking clients. The role focuses on building trustworthy, source‑cited AI assistants that meet strict compliance requirements.

Key responsibilities

• Design and implement end‑to‑end RAG pipelines: document chunking, embeddings, vector search, re‑ranking, prompting and grounding.
• Develop and maintain prompt designs, guardrails and evaluation harnesses.
• Fine‑tune or domain‑adapt large language models as needed.
• Ensure every AI answer is traceable to approved source documents and complies with regulatory standards.
• Implement content safety, data masking, retrieval‑time access control and audit trails.

Required profile

• 4+ years of experience in ML/NLP or software engineering.
• At least 1.5 years of hands‑on experience building LLM/RAG applications.
• Proven delivery of a complete RAG system including ingestion, embeddings, vector search and evaluation.
• Experience with hallucination control, grounding, citation of sources and structured quality evaluation.
• Familiarity with fine‑tuning, prompt‑injection defense and jailbreak mitigation.

Required skills

• Azure OpenAI Service, Azure AI Search, Azure AI Content Safety, Azure AI Foundry (prompt flow/evaluation), Copilot Studio.
• Python with LangChain, LlamaIndex, Semantic Kernel.
• Hugging Face Transformers and open embedding models.
• Vector databases: pgvector, Qdrant, Chroma, Milvus.
• Evaluation tools: RAGAS, DeepEval.
• Local serving: Ollama, vLLM.
• Fine‑tunable models: Llama, Mistral with LoRA/PEFT.
• Guardrail frameworks: Llama Guard, NeMo Guardrails.
• Access control integration with Entra ID and data masking for PII/PCI.