Job title: Senior Agentic AI Engineer
Experience : 6+ Years
Location : Pune/Banglore
Work Mode : Remote
Notice Period: Immediate

The Opportunity
This is a role in a new AI engineering squad. You'll work across the full agentic stack from designing agent architectures and RAG pipelines to shipping autonomous agents that replace manual processes in production. Hands-on, delivery-focused, and consequential from day one.

What You'll Do
Agents
• Design and build multi-agent systems for autonomous reasoning, planning, and task execution
• Architect orchestration patterns; tool use, memory, reflection loops, human-in-the-loop handoffs
• Deliver across use cases: developer productivity agents, knowledge retrieval agents, and process
automation
• Favour GCP-native and enterprise-grade managed services before writing custom infrastructure, we're
not in the business of rebuilding what a major cloud vendor will productise in six months
RAG, Knowledge & Graph
• Design and implement end-to-end RAG architectures ingestion, chunking, embedding, retrieval,
reranking
• Select and manage vector databases appropriate to the use case
• Build knowledge pipelines that turn unstructured client data into queryable intelligence
• Apply knowledge graph structures (e.g. Neo4j) where relational or semantic reasoning adds value
beyond flat vector retrieval
Evaluation & Observability
• Design and own eval frameworks; automated evals, LLM-as-a-Judge, task-completion harnesses
• Define what 'working correctly' looks like for each agent and measure it continuously in production




• Instrument agents for cost, latency, and quality
Platform & Collaboration
• Contribute to platform standards how agents are built, deployed, monitored, and governed
• Ensure agents are production-grade: observable, cost-aware, GDPR and EU AI Act compliant
• Work with engineering, business stakeholders, and the Director of Platform, Engineering & AI to
translate process problems into agentic solutions
What We're Looking For
Must-Have
• Production experience building and deploying agentic AI systems not just prototypes
• Strong Python engineering: clean, tested, maintainable code
• Hands-on RAG pipeline design chunking strategies, embedding models, retrieval, evaluation
• Production experience with at least one vector database
• LLM evaluation experience automated evals, LLM-as-a-Judge, or equivalent agent accuracy
measurement
• Familiarity with agent frameworks (LangChain, LangGraph, CrewAI, AutoGen, or similar) and the
judgement to know when not to use them
• Solid LLM fundamentals: prompting, tool calling, structured outputs, context management
• Cloud infrastructure comfort GCP preferred; strong AWS/Azure considered
Nice-to-Have
• Google Agent Development Kit (ADK)



or Vertex AI Agent Builder
• MCP (Model Context Protocol) and agent-to-agent communication patterns
• Knowledge graphs or graph databases (Neo4j, JanusGraph) in retrieval or reasoning contexts
• LLMOps / observability tooling LangSmith, Langfuse, or equivalent
• Experience building agents in regulated or enterprise environments
• Background in BPO, AP automation, or document processing workflows
What Good Looks Like
• Shipped agents, not demos production deployment matters more than framework familiarity
• Engineering mindset you care about reliability, cost, testability, and observability
• Built evals, not vibes you measure whether agents work and instrument from day one
• Build-vs-buy judgment you reach for GCP-native and enterprise tooling before writing custom
infrastructure you've seen bespoke orchestration layers made redundant by a platform release, and
you build accordingly
• Comfortable with ambiguity you help define the patterns rather than follow them
• Business curious you want to understand the process you're automating, not just the stack
A Note on This Role
Agentic AI engineering is genuinely new. We don't expect 10 years of experience in a discipline that's 2 years
old. What we do expect is evidence of building real things, robust software engineering foundations, and the
intellectual curiosity to stay ahead of a fast-moving field. If you've been doing this seriously for 18 months and
have production systems to show for it, we want to hear from you.

📌 Senior Agentic AI Engineer (Pune)
🏢 Optimum Data Analytics
📍 Pune