Associate Developer, Agentic AI

Job Summary

Associate Developer, Agentic AI

We are seeking an Associate Developer Agentic AI with 1-2 years of experience who already operates well beyond that level. Youve shipped agents to production, debugged them when they broke, and built integrations that survived real load. You dont wait to be told what to do - you identify the problem, make a call, and close the loop.

Responsibilities

• Design, develop, and own LLM-powered AI agents and RAG-based features end-to-end - from architecture through production - using Python (FastAPI), ReactJS, and frameworks like LangGraph.
• Architect and implement multi-agent systems with orchestrator/specialist patterns, managing agent state, memory, and inter-agent communication at scale.
• Lead integration of OAuth-based third-party connectors (ERP, CRM, productivity, supplier systems) via REST APIs and MCP (Model Context Protocol), with production-grade auth, retry, and error handling.
• Own data layer decisions across PostgreSQL, MongoDB, Redis, and vector databases - optimizing for agent memory, retrieval, and scale.
• Design and operate containerized agent services on GCP (GKE, Cloud Run, Pub/Sub, Vertex AI) - including deployment pipelines, observability, and incident response.
• Implement human-in-the-loop (HITL) patterns - approval gates, edge case escalation, confidence thresholds - within long-running agentic workflows.
• Drive LLM evaluation and observability using LangSmith, RAGAS, or DeepEval - setting quality bars, not just running evals.
• Contribute meaningfully to architectural discussions and code reviews; raise the quality bar for the team.
• Partner with product and senior architects to scope and decompose enterprise procurement use cases into shippable agent workflows.

Position Requirements

Must-Have

• 1-2 years of engineering experience - what matters is that at least a significant portion of it involved building and shipping LLM-powered systems to production, not just experimenting.
• Masters or Bachelors in Computer Science or related discipline; IIT, NIT, IIIT, BITS or equivalent strongly preferred.
• Strong production experience in Python (FastAPI) and ReactJS.
• Hands-on experience with LangGraph - stateful, multi-step agent workflows, not just tutorials.
• Demonstrated ability to design multi-agent orchestration - task decomposition, specialist agents, state handoff, failure recovery.
• Solid experience with PostgreSQL, MongoDB/Firestore, Redis, and at least one vector database; familiarity with hybrid search and re-ranking.
• Production experience with Docker, Kubernetes, and GCP or equivalent cloud.
• Strong grasp of OAuth 2.0, JWT, and secure API integration patterns.
• Experience with LLM evaluation and observability tooling (LangSmith, RAGAS, DeepEval, or Phoenix).

Strong Plus

• Familiarity with MCP and/or A2A protocol for agent interoperability.
• Experience with managed LLM services - Vertex AI, AWS Bedrock, or Azure OpenAI - in production.
• Exposure to event streaming (Kafka, Pub/Sub) in agentic or data pipeline contexts.
• Open-source contributions or shipped personal projects in the agent/LLM space - a strong signal here carries more weight than years of experience elsewhere.

Who You Are

• Youve already built things most engineers with your experience havent - and you have the GitHub history, shipped product, or war stories to prove it.
• You own outcomes, not just tasks - you track a problem from requirement to production without being chased.
• Youve debugged agents that hallucinated, looped, or failed silently and know how to build guardrails.
• You make pragmatic calls - you know when to reach for a framework and when to write plain Python.
• You communicate with clarity and make the people around you better without needing a title to do it.

Disclaimer: This job posting has been aggregated from external source. Role details, content, and availability are subject to change. Applicants are advised to confirm the latest information directly on the company website before applying.