AI / LLM Engineer

Infosys; Pune District; Presencial; INR 2.400.000 - 4.800.000; Jornada completa

Descripción de la vacante

Infosys in Pune invites senior AI/ML engineer to shape enterprise-grade LLM solutions. You will design and implement production-ready AI components, build retrieval-augmented generation pipelines, and collaborate with backend and DevOps teams to containerize services.

The role emphasizes governance, safety, and explainability, with integration to Azure OpenAI, OpenAI, Google Gemini, and AWS Bedrock. You will develop reusable prompts, evaluation routines, and oversee end-to-end AI workflows in

Formación

• 10+ years in software engineering or AI/ML roles.
• Strong hands-on Python programming experience.
• Experience with RAG, vector databases, embeddings, prompts, evaluation.
• Familiar with LangChain, LangGraph, LlamaIndex, Semantic Kernel, AutoGen, CrewAI.
• REST APIs, microservices, SQL, Git, testing and software practices.
• Knowledge of document extraction, semantic search, NLP, hallucination risk and prompt safety.
• Able to build production-oriented LLM workflows and AI services.

Conocimientos

Python programming

LLM-based application development

RAG and vector databases

Embeddings and prompt engineering

Evaluation frameworks

AI service integration

REST APIs and microservices

NLP, document extraction, semantic /pr

Production-oriented AI components

AI governance and safety practices

Herramientas

LangChain

LangGraph

LlamaIndex

Semantic Kernel

AutoGen

CrewAI

Azure OpenAI

Azure AI Foundry

OpenAI APIs

Google Gemini

Hugging Face

RAGAS

DeepEval

Promptfoo

LangSmith

Descripción del empleo

Minimum 10+ years of experience in software engineering, AI/ML engineering, applied ML, data science engineering or related roles.

• Strong hands-on Python programming experience and practical exposure to LLM-based application development.
• Experience with RAG, vector databases, embeddings, prompt engineering, evaluation frameworks and AI service integration.
• Experience with frameworks such as LangChain, LangGraph, LlamaIndex, Semantic Kernel, AutoGen, CrewAI or equivalent tools.
• Working knowledge of REST APIs, microservices, SQL, structured data concepts, Git workflows, testing and software engineering practices.
• Understanding of document extraction, semantic search, NLP, retrieval quality, hallucination risk, prompt safety and AI evaluation methods.
• Ability to build production-oriented AI components rather than isolated proof-of-concept demos. Design and implement LLM-powered workflows for summarization, narrative generation, classification, extraction, contextual reasoning, explanation and reviewer-assist use cases.
• Build retrieval-augmented generation pipelines including document ingestion, chunking, embedding generation, metadata tagging, vector indexing, retrieval tuning and grounded response generation.
• Develop reusable prompt templates, prompt versions, context builders, response schemas, evaluation routines and AI orchestration services.
• Integrate with enterprise AI services such as Azure OpenAI, Azure AI Foundry, OpenAI APIs, Google Gemini, Anthropic, Hugging Face or equivalent approved platforms.
• Implement AI run logging, prompt/model metadata capture, evidence citations, output traceability, reviewer feedback capture and human-in-the-loop controls.
• Build AI evaluation routines for answer quality, retrieval quality, hallucination checks, regression testing, consistency and groundedness.
• Collaborate with backend and DevOps teams to containerize AI services, deploy them securely, monitor usage, track costs and troubleshoot production issues.
• Support responsible AI practices such as prompt injection checks, data leakage prevention, policy-based guardrails and AI output validation. Experience with Azure OpenAI, Azure AI Search, Azure Document Intelligence, Google AI Studio/Gemini, AWS Bedrock or Vertex AI.
• Exposure to RAG evaluation tools such as RAGAS, DeepEval, Promptfoo, LangSmith or equivalent frameworks.
• Experience with AI governance, prompt/model registry, AI audit logs, explainability, groundedness checks and human review workflows.