Agentic AI Engineer / AI-ML Engineer

Client-calibrated JD for sourcing candidates with real Agentic AI implementation experience

Experience 5-12 Years

Location Pune / Hybrid

Role Type Hands-on Agentic AI / AI-ML Engineering

Important Client Feedback

Profiles with only GenAI, RAG, Data Science, ML, NLP, or chatbot experience are not sufficient. Client expects demonstrable Agentic AI, multi-agent, tool-calling and orchestration experience.

Role Summary

We are looking for a hands-on Agentic AI Engineer who has built and deployed production-grade AI Agents and

Multi-Agent Systems using modern Agentic AI frameworks. The ideal candidate should have strong Python development skills along with practical experience designing autonomous AI workflows, orchestrating multiple agents, implementing tool-calling capabilities, integrating enterprise systems, and building scalable RAG-based AI solutions.

Mandatory Skills

Skill Area Required Capability

Core AI & Python

Solid Python development, Machine Learning fundamentals,

LLMs such as GPT/Claude/Gemini/Llama/Mistral, Generative

AI, Prompt Engineering.

Agentic AI - Mandatory

Hands-on Multi-Agent Systems, Agent Orchestration, Tool

Calling / Function Calling, Agent Memory, Planning &

Reasoning Workflows, Human-in-the-Loop workflows and AI-

driven workflow automation.

Agentic Frameworks Practical experience with one or more: LangGraph, CrewAI,

AutoGen, Semantic Kernel, OpenAI Agents SDK.

RAG & Retrieval

Production-grade RAG pipelines, LangChain or LlamaIndex,

embeddings, retrieval optimization, re-ranking, chunking strategies and citation frameworks.

Vector Databases Hands-on experience with one or more: Pinecone,

ChromaDB, Weaviate, FAISS, Qdrant or pgVector.

Backend Engineering FastAPI, REST APIs, microservices, API integrations,





authentication and authorization.

Cloud Platforms Experience with at least one: Azure OpenAI, AWS Bedrock or

GCP Vertex AI.

Preferred Skills

 MCP (Model Context Protocol)

 LLMOps / AgentOps

 LangSmith or similar tracing/observability tools

 Prompt Flow or equivalent AI workflow tooling

 Agent monitoring, evaluation, guardrails and governance

 AI security and responsible AI practices

 Docker, Kubernetes and CI/CD pipelines

Must-Have Project Experience

 Designed and implemented at least 2-3 production or enterprise-grade Agentic AI use cases.

Agentic AI Engineer JD | Client-Calibrated Sourcing Version

 Built multi-agent workflow designs and agent-to-agent communication flows.

 Implemented tool-calling or function-calling with APIs, databases, enterprise systems or knowledge sources.

 Built or optimized enterprise RAG systems with retrieval quality improvements.

 Implemented autonomous reasoning workflows, planning, memory or state management.

 Added AI agent observability, monitoring, evaluation or quality checks.

 Developed AI copilots or workflow automation using AI agents.

Mandatory Screening Questions

No. Screening Question Expected Evidence

1 Which Agentic AI framework has the candidate used?

LangGraph, CrewAI, AutoGen,

Semantic Kernel, OpenAI Agents SDK,

etc.

2 Has the candidate built a Multi-Agent

System?

Clear project example,



not only theoretical knowledge.

3 Has the candidate implemented Tool

Calling / Function Calling?

APIs, database actions, ticketing workflows, enterprise systems, etc.

4 Has the candidate built a production

RAG pipeline?

Chunking, embeddings, vector DB,

retrieval optimization and evaluation.

5 Which Vector Database has been used? Pinecone, ChromaDB, Weaviate,

FAISS, Qdrant, pgVector, etc.

6 Which LLM platform has been used? OpenAI, Azure OpenAI, AWS Bedrock,

Vertex AI, Claude, Gemini, etc.

7 Has the candidate deployed AI Agents into production?

Deployment details, users, workflows,

monitoring or support.

8 Has the candidate implemented Agent

Memory and Reasoning workflows?

Memory/state management, planning,

reflection or routing logic.

Reject If

 Only Data Scientist profile without agentic implementation.

 Only NLP or classical ML profile.

 Only Prompt Engineering experience.

 Only basic LangChain chatbot experience.

 No Multi-Agent implementation.

 No Tool Calling / Function Calling experience.

 No Agent Orchestration experience.

 No production-grade RAG or enterprise integration exposure.

Keywords for Search

Agentic AI, Multi-Agent Systems, LangGraph, CrewAI, AutoGen, Semantic Kernel, OpenAI Agents, Tool

Calling, Function Calling, RAG, LlamaIndex, LangChain, MCP, Vector Database, Pinecone, ChromaDB,

Weaviate, Qdrant, FAISS, pgVector, FastAPI, Vertex AI, Bedrock, Azure OpenAI, AI Copilot, Autonomous

Agents, AI Orchestration, AgentOps, LLMOps.

Intake Note

Do not shortlist profiles based only on GenAI/RAG/Chatbot keywords. TA should validate actual Agentic AI project evidence: multi-agent workflows, tool-calling, orchestration, memory, reasoning, RAG

📌 AI ML Python Lead (Mumbai)
🏢 Wipro
📍 Mumbai