Agentic AI & GenAI Engineer Intern

**Agentic AI \& GenAI Engineer Intern** AI Platform · SDLC Intelligence · Multi-Agent Systems Internship \| 6–12 months \| 6\+ months experience required **What We're Building** We're building a fully agentic, LLM-native platform that rewires how software gets designed, tested, and shipped — from requirements to release. Think autonomous agents that read tickets, generate code, catch defects, and trigger CI/CD pipelines — all orchestrated through a system you help design. This isn't a "run some notebooks" internship. You'll work hands-on with multi-agent frameworks, MCP servers, RAG pipelines, and LLM tool-use patterns — and your work ships into real systems. If you've already spent 6\+ months tinkering with LLMs, building backend services, or wiring up agentic workflows, you'll hit the ground running. **What You'll Actually Do** Agentic Systems * Build and iterate on multi-agent orchestration flows using LangChain, LangGraph, or AutoGen * Design and expose MCP (Model Context Protocol) servers to give agents structured access to tools, repos, and APIs * Implement agent memory, reflection loops, and tool-use patterns (ReAct, Plan-and-Execute) LLM \& GenAI Engineering * Work on LLM use cases across the SDLC: requirement parsing, test case generation, code review, defect triage * Build and tune RAG pipelines over codebases, logs, and defect databases using vector stores * Experiment with prompt strategies — chain-of-thought, few-shot, structured output, tool-calling Backend \& API Integration * Build FastAPI/REST backend services that expose agent capabilities to product surfaces * Connect LLM workflows into CI/CD pipelines (GitHub Actions, Jenkins) for automated code analysis and testing * Integrate with coding tools — IDE plugins, PR bots, CLI assistants Evaluation \& Reliability * Design evaluation frameworks for agent outputs — accuracy, hallucination rate, tool-call fidelity * Build observability into pipelines: tracing, structured logging, drift detection * Own experiments end-to-end: hypothesis → implementation → measurement → ship **What You Need (Must-Have)** * Python 3\.x — solid, not just scripting * Hands-on experience with LLMs and prompt engineering * Understanding of RAG pipelines and vector search * REST API development and integration * Git and version control * Basic ML concepts (classification, evaluation metrics) * Familiarity with agentic workflows and tool-use patterns **Strong Advantage** * LangChain, LangGraph, or similar orchestration frameworks * MCP server design and integration * Vector databases (Chroma, Pinecone, Qdrant) * FastAPI or Flask for backend services * CI/CD pipelines (GitHub Actions, Jenkins) * Cloud platforms (Azure, AWS, or GCP) * Docker basics * TypeScript fundamentals **Bonus Points** * AutoGen or CrewAI experience * LLM fine-tuning (LoRA, PEFT) * AI coding tools (Cursor, Copilot, Cline) * SDLC or QA process exposure * OpenTelemetry or observability tooling **The Baseline We Expect** 6\+ months of hands-on experience with at least one of: LLM application development, backend API engineering, or agentic/GenAI systems. This could be from a previous internship, freelance project, open-source contribution, or a personal project you can walk us through. We care about what you've actually built — not where you studied. Pay: ₹20,000\.00 - ₹40,000\.00 per month Work Location: In person