GenAI Engineer (LLM / RAG / Multimodal)

Work Location : Chennai

Work Mode : Hybrid

Notice: Immediate to 30 days

Experience : 3 - 7 years

Role Summary

• The GenAI Engineer designs, builds, and optimizes LLM-powered features such as chatbots, copilots, summarizers, and document assistants.
• You will work with LLM APIs, vector search, and orchestration frameworks to turn business workflows into reliable, scalable GenAI solutions.

Key Responsibilities

AI Feature Development

• Build GenAI use cases: Q&A, summarization, classification, content generation, conversational assistants, copilots.
• Use LLM APIs (OpenAI, Azure OpenAI, etc.) to implement production-ready features.
• Implement guardrails and safety filters to reduce hallucinations and unsafe outputs.

RAG & Retrieval Pipelines

• Implement RAG pipelines by combining LLMs with vector and hybrid search.
• Collaborate with Data Engineers on chunking, embedding selection, and retrieval strategies.
• Evaluate retrieval and answer quality and iterate based on metrics and user feedback.

Orchestration & Agentic Flows

• Orchestrate multi-step workflows using frameworks like LangChain, Semantic Kernel, LangGraph, Microsoft Agent Framework, Copilot Studio, or similar.
• Implement tool/function calling to let LLMs interact with APIs, databases, and business systems.
• Design agent workflows for tasks like research, summarization, routing, and workflow automation.

Preferred candidate profile

• Bachelors/ Master’s degree in computer science, Engineering, Data Science, or equivalent experience.
• 3+ years in software or ML engineering, with 1–2 years in LLM / GenAI projects.
• Strong programming skills in Python (preferred) or Node.js/C#.
• Hands-on experience with: LLM APIs (OpenAI / Azure OpenAI / similar).
• At least one orchestration framework: LangChain, Semantic Kernel, LangGraph, Microsoft Agent Framework, Copilot Studio, etc.
• Building backend services (FastAPI, Flask, Node, .NET, etc.).
• Understanding of: RAG, embeddings, vector search, prompt engineering.
• JSON-based tool/function calling and schema validation.
• Basic security and privacy practices for AI applications.