Gen AI Lead

Job description

**AI Engineer** **Experience: 6 – 9 years** **About the role** We’re looking for a hands‑on Senior AI Engineer who is passionate about building production‑grade Retrieval‑Augmented Generation (RAG) systems and LLM‑powered agents and applications on the Azure stack. You’ll design, ship, and operate AI services end‑to‑end—grounding models with trustworthy data, instrumenting quality and safety, and automating reliable cloud infrastructure with Terraform. This is a role for a self‑starter with a builder’s mindset, high ownership, and a deep curiosity for how modern AI can solve real business problems. **What you’ll do** - Design and ship RAG systems end‑to‑end: data connectors, indexing pipelines, hybrid search (vector), chunking/metadata strategy, citation/attribution, and evaluation at scale. - Build applications with Python and open‑source frameworks (e.g., LangChain, LlamaIndex, FastAPI), GHCP, Claude , etc leveraging prompt engineering, tools/agents, and structured outputs. - Operate on Azure: Azure OpenAI / Models, Azure AI Search, Azure ML, AKS, Functions, Key Vault, Event Hub/Service Bus, App Insights; implement auth, secrets, and network isolation. - Fine‑tune and optimize LLMs (LoRA/QLoRA, adapters, domain instruction‑tuning), manage embeddings and vector stores (Azure AI Search etc.), and evaluate model quality. - Productionize: CI/CD, containerization (Docker/Kubernetes), observability (OpenTelemetry/App Insights), tracing, cost controls, feature flags, and progressive delivery. - Infrastructure as Code: model and provision Azure resources with Terraform, codify environments (dev/test/prod), enforce policies, and bake in security from day one. - Quality, safety, and governance: implement automated evaluations (accuracy, hallucination, groundedness), bias/fairness checks, human‑in‑the‑loop gates, prompt/change management, drift monitoring, and incident response aligned to our Model Risk Management and AI governance practices (hallucination testing, bias/fairness, HITL, prompt management, monitoring & version control). **What makes you a great fit** - Aptitude & attitude: You learn fast, simplify complex problems, and default to action. You take ownership, communicate clearly, and raise the bar for engineering craft. - Deep passion for AI: You keep up with the state of the art and translate research into pragmatic, reliable solutions. **Qualifications** **Required** - 5+ years of software engineering with Python in production (APIs, services, testing, packaging). - Hands‑on RAG experience: embeddings, retrieval strategies, chunking, metadata/routing, evals, and guardrails. - Open‑source frameworks: strong with LangChain (or similar), plus experience with FastAPI/Flask and async patterns. - Azure: practical experience with Azure OpenAI/Models, Azure AI Search, Azure ML, AKS/Container Apps, Key Vault, App Insights/Log Analytics, and Hybrid Private Networking. - Terraform: modules, CI/CD integration, and handling nested data structures. - MLOps/DevOps: Docker/Kubernetes, CI/CD (Gitlab or Azure DevOps), secrets management, and automated testing. - Solid understanding of LLMs (prompting, function/tool calling, structured outputs, rate limiting, token/cost management). **Preferred** - LLM fine‑tuning (LoRA/QLoRA, PEFT), dataset curation/red‑teaming, distillation, or adapters; evaluation frameworks (RAGAS, pairwise/criteria‑based). - Vector databases - Security & compliance for regulated industries (PII handling, isolation patterns, policy as code). - Typescript/Node for front‑ends or extensions (e.g., plugins, Copilot‑style integrations). - **Observability** : Langfuse/ OpenTelemetry traces for prompts/retrieval, eval dashboards, automated rollback criteria.