Headline

Lead Agentic AI Engineer | Building Enterprise Agentic AI Systems with LangGraph, RAG, MCP, Observability, Guardrails & Evaluation Frameworks

About

I help design and build reliable Agentic AI systems for enterprise use cases, with a strong focus on RAG, LangGraph-based orchestration, MCP tool integration, guardrails, AI evaluation, and production-readiness. My current focus is on building end-to-end Agentic AI architectures that go beyond simple chatbot implementations. I work on systems that can understand user intent, retrieve contextual knowledge, use tools safely, apply input/output guardrails, evaluate retrieval quality, validate responses using LLM-as-a-Judge, and provide traceable, explainable outputs. Key areas I work on include: • Agentic AI workflow design using LangGraph • Retrieval-Augmented Generation systems and vector databases • MCP-based tool and system integration • Input and output guardrail design • Search evaluation and retrieval evaluation • Answer evaluation using LLM-as-a-Judge • RAG quality improvement and hallucination reduction • Multi-agent orchestration and decision workflows • Human-in-the-loop validation for enterprise AI systems • AI governance, traceability, observability and reliability frameworks I am particularly interested in building production-grade Agentic AI platforms where reliability, safety, evaluation, and governance are treated as core engineering requirements - not afterthoughts. My background in enterprise automation, workflow systems, API-based platforms, and process orchestration gives me a strong foundation to build AI systems that are practical, measurable, and ready for real business adoption.

Location

Pune Division

AI stack

claude code

claude

codex

gemini

langchain

langgraph

langsmith

mcp

ollama

openai

railway

claudecode

githubcopilot

Guardrails AI

Work experience

Lead Agentic AI Engineer at Jaguar Land Rover India, May 2025 - Present · 1 yr 3 mos, • Designed Agentic AI system architectures using LangGraph for multi-step reasoning, routing, orchestration, and tool-based workflows. • Built RAG-based knowledge systems, vector retrieval, structured prompts, and contextual response generation. • Implemented input guardrails to validate user intent, detect unsafe prompts, prevent prompt injection, and enforce domain boundaries. • Designed output guardrails to validate generated responses for relevance, safety, hallucination risk, tone, and policy alignment. • Integrated MCP-style tool interaction patterns to enable agents to safely interact with external systems and business workflows. • Developed evaluation layers including search evaluation, retrieval evaluation, answer evaluation, and LLM-as-a-Judge based scoring. • Created traceability mechanisms to observe agent decisions, retrieved sources, intermediate reasoning steps, and final outputs. • Worked on enterprise AI use cases involving workflow automation, document understanding, knowledge retrieval, and decision support. • Defined reusable Agentic AI patterns for human-in-the-loop review, governance checkpoints, and safe autonomous execution. • Built an observability layer for Agentic AI workflows to trace LangGraph executions, monitor retrieval quality, inspect tool calls, evaluate answer quality, and capture guardrail decisions.

AI Automation Lead at FIS, Jul 2023 - May 2025 · 1 yr 11 mos, Pune District, • Led enterprise automation initiatives across API-driven and event-driven banking platform components. • Worked on distributed event-flow architecture involving outbox patterns, active/archive event tables, Kafka topics, webhook engines, downstream consumers, and audit trails. • Designed automation coverage for complex asynchronous workflows, including event generation, event publishing, webhook delivery, retry behavior, failure scenarios, and DLQ processing. • Built strong expertise in API orchestration, message-driven systems, system contracts, integration workflows, and enterprise platform reliability. • Worked with authentication and security patterns including Basic Auth, OAuth, mTLS, and HMAC-based integrations across webhook-driven platforms. • Supported cloud-native automation workflows involving Kubernetes/OpenShift, containerized execution, CI/CD pipelines, and environment-level validation. • Developed reusable automation patterns for validating platform behavior across multiple tenants, consumers, event types, topics, and endpoint configurations. • Created reliability-focused automation for failure-mode validation, retry scenarios, endpoint failures, authorization failures, duplicate event handling, and message delivery assurance. • Gained practical experience in system observability through logs, audit tables, message traces, event status tracking, and end-to-end workflow validation. • Built a strong foundation for AI Automation and Agentic AI engineering by working on API-based actions, event-driven orchestration, workflow reliability, guardrail-style validations, and enterprise integration patterns.

Intelligent Automation Developer - RPA, Data Science & AI/ML at Tata Consultancy Services, Jun 2019 - Jul 2023 · 4 yrs 2 mos, Pune Division, • Built early foundation in Intelligent Automation through RPA development, process automation, and AI/ML-based predictive analytics. • Worked with UiPath and Blue Prism to automate business workflows involving UI actions, rule-based decisions, exception paths, and process handoffs. • Developed strong understanding of how enterprise workflows can be digitized, automated, monitored, and improved through automation technologies. • Contributed to an AI/ML-based predictive analytics use case for assembly line fault prediction, applying data-driven techniques to support operational decision-making. • Gained practical experience in workflow automation, bot orchestration, exception handling, operational logging, and process reliability.

Consultant at Capgemini, Feb 2016 - Apr 2019 · 3 yrs 3 mos, Pune Division, • Worked on Cognitive RPA solutions using WorkFusion for document-centric business process automation. • Used OCR capabilities to extract information from scanned documents, forms, PDFs, and image-based inputs. • Supported automation workflows involving document ingestion, OCR extraction, field validation, business rule checks, and downstream processing. • Worked on cognitive automation patterns where RPA was combined with OCR and intelligent data extraction to reduce manual document handling. • Built understanding of semi-structured and unstructured document processing, exception handling, confidence scoring, and human review workflows. • Gained practical experience in Intelligent Document Processing, cognitive RPA, and AI-assisted automation use cases.

Performance Engineer at Atos Syntel, Jun 2013 - Aug 2015 · 2 yrs 3 mos, Pune Division, • Successfully analysed requirements and prepared a Performance Test Plan, test scenarios, and test data • Developed virtual user scripts for various business processes using HP Load Runner VUGen, and enhanced the scripts by adding transactions, parameterising constant values, and correlating dynamic values • Collaborated with business analysts to analyse business scenarios, as well as ensure timely and high-quality delivery • Independently created load and stress scenarios for performance testing using the Load Runner Controller, improved Load Runner scripts for testing new builds of the application, and uploaded and executed the scripts using Performance Centre • Utilised Load Runner to evaluate the performance of various reports and queries under load, analysed the results using the Load Runner Analysis tool, and identified bottlenecks by analysing online monitors and graphs • Effectively extracted reports from the Analysis Tool and consolidated them to prepare a final summary report • Mentored new team members on Load Runner and Performance Centre and participated in performance meetings with developers and clients as the offshore representative

Education

Bachelor of Computer Science, Computer Science, Department of Computer Science, Pune University

SSPMS Junior College