Experience
4 - 8 yrs
Job Location
Pune, India
Vacancy
1
Designation
Senior Artificial Intelligence Engineer
Job Type
ONSITE
Job Description
Title: Senior Specialist Ai Developer
Location: Pune
Exp: 4-8 Years
Key Responsibilities
Build and maintain the LangGraph orchestration engine: state graph definition, parallel fan-out, conditional routing, checkpointing, retry and fallback logic.
Develop and deploy specialized AI agent nodes Gate Readiness Agent, Quality Check Agent, Search RAG Agent, Notification Agent, Action Execution Agent within the multi-agent architecture.
Implement and manage MCP (Model Context Protocol) server pods: SNPD Core MCP, SAP MCP, PLM MCP, Doc Search MCP ensuring RBAC enforcement, cost masking, and audit logging at each layer.
Integrate Azure OpenAI endpoints (GPT-4o for reasoning, GPT-4o-mini for intent classification, text-embedding-3-large for RAG) via Azure APIM.
Build the RAG pipeline: chunking strategy, embedding generation, Azure AI Search vector index management, retrieval tuning, and citation-accurate response synthesis.
Implement the Conversational Copilot API including streaming responses, session state management, RBAC-scoped tool visibility, and Human-in-the-loop confirm nodes for write operations.
Maintain and version the Prompt Registry (system prompts, few-shot examples, output schemas per AI feature) with admin-configurable deployment no code release required.
Implement the AI Audit Logger: every query, model version, tool calls, and response logged to audit_log for IATF 16949 compliance.
Develop AI microservices APIs (REST / event-driven) that expose AI capabilities to the application layer; integrate with .NET Core backend services.
Contribute to OWASP AI security practices: prompt injection defence, output validation, hallucination guardrails, and content safety filters.
Write unit and integration tests for agent nodes, MCP tools, and RAG retrieval quality using Promptfoo / pytest frameworks.
Participate in code reviews, architecture decision records, and sprint ceremonies as part of the AI squad.
Technical Skills required
Python 3.10+ (primary language) FastAPI / REST API design LangGraph / LangChain state graph, tool use Docker / Azure AKS (Kubernetes) Azure OpenAI (GPT-4o, GPT-4o-mini, embeddings) Azure API Management (APIM) Azure AI Search (vector + keyword hybrid) Azure AI Foundry / Azure ML MCP (Model Context Protocol) server design Prompt engineering few-shot design RAG architecture chunking, retrieval, reranking Git, CI/CD (Azure DevOps / GitHub Actions)
Good To Have
Experience with agentic AI frameworks: CrewAI, AutoGen, Semantic Kernel.
Familiarity with SAP S/4 HANA OData APIs or Teamcenter PLM REST APIs.
Knowledge of IATF 16949 quality management or automotive NPD processes.
Experience with LLM evaluation frameworks: RAGAS, Promptfoo, TruLens.
Hindi / Marathi language NLP supporting regional user queries.
", 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.
Location: Pune
Exp: 4-8 Years
Key Responsibilities
Build and maintain the LangGraph orchestration engine: state graph definition, parallel fan-out, conditional routing, checkpointing, retry and fallback logic.
Develop and deploy specialized AI agent nodes Gate Readiness Agent, Quality Check Agent, Search RAG Agent, Notification Agent, Action Execution Agent within the multi-agent architecture.
Implement and manage MCP (Model Context Protocol) server pods: SNPD Core MCP, SAP MCP, PLM MCP, Doc Search MCP ensuring RBAC enforcement, cost masking, and audit logging at each layer.
Integrate Azure OpenAI endpoints (GPT-4o for reasoning, GPT-4o-mini for intent classification, text-embedding-3-large for RAG) via Azure APIM.
Build the RAG pipeline: chunking strategy, embedding generation, Azure AI Search vector index management, retrieval tuning, and citation-accurate response synthesis.
Implement the Conversational Copilot API including streaming responses, session state management, RBAC-scoped tool visibility, and Human-in-the-loop confirm nodes for write operations.
Maintain and version the Prompt Registry (system prompts, few-shot examples, output schemas per AI feature) with admin-configurable deployment no code release required.
Implement the AI Audit Logger: every query, model version, tool calls, and response logged to audit_log for IATF 16949 compliance.
Develop AI microservices APIs (REST / event-driven) that expose AI capabilities to the application layer; integrate with .NET Core backend services.
Contribute to OWASP AI security practices: prompt injection defence, output validation, hallucination guardrails, and content safety filters.
Write unit and integration tests for agent nodes, MCP tools, and RAG retrieval quality using Promptfoo / pytest frameworks.
Participate in code reviews, architecture decision records, and sprint ceremonies as part of the AI squad.
Technical Skills required
Python 3.10+ (primary language) FastAPI / REST API design LangGraph / LangChain state graph, tool use Docker / Azure AKS (Kubernetes) Azure OpenAI (GPT-4o, GPT-4o-mini, embeddings) Azure API Management (APIM) Azure AI Search (vector + keyword hybrid) Azure AI Foundry / Azure ML MCP (Model Context Protocol) server design Prompt engineering few-shot design RAG architecture chunking, retrieval, reranking Git, CI/CD (Azure DevOps / GitHub Actions)
Good To Have
Experience with agentic AI frameworks: CrewAI, AutoGen, Semantic Kernel.
Familiarity with SAP S/4 HANA OData APIs or Teamcenter PLM REST APIs.
Knowledge of IATF 16949 quality management or automotive NPD processes.
Experience with LLM evaluation frameworks: RAGAS, Promptfoo, TruLens.
Hindi / Marathi language NLP supporting regional user queries.
", 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.
No Referrers Available
There are currently no referrers available for this job. You can still apply, will let you know once there is any referrer available.