Headline

Building GenAI Solutions | Agentic Frameworks | LLM Apps on IBM Watsonx | AI Automation Engineer

About

Generative AI Engineer with 3+ years of experience designing and deploying agentic systems, LLM-powered pipelines, and retrieval-augmented generation (RAG) solutions in production. Skilled in orchestrating multi-agent frameworks using LangChain, LangGraph, and custom planning agents. Experienced with embedding models, vector databases (Qdrant, Milvus), rerankers, and OpenAI APIs. Strong foundation in NLP, prompt engineering, and scalable backend development using FastAPI. Currently building autonomous agents that can understand user ML/NLP needs, gather data, build and evaluate models, and retrain based on feedback. Passionate about developing intelligent, self-improving systems for real-world impact.

Location

Hyderabad

AI stack

aistudio

codex

kilocode

meta

Work experience

AI Engineer at Bowmen Group, Jul 2024 - Present · 2 yrs, Remote, Driving AI innovation at Bowmen Group as an AI Engineer. Currently leading the development and deployment of advanced multi-agent systems, achieving 80-90% accuracy through strategic LLM orchestration (including IBM Granite, Llama, Mistral). Focused on building scalable, secure, and high-performing AI solutions that deliver tangible business value. Key responsibilities include architecting custom agent frameworks, integrating real-time knowledge retrieval, and designing dynamic user interfaces.

Artificial Intelligence Engineer at Bowmen Group, Jul 2024 - Present · 2 yrs, Bowmen OS — Agentic GRC Automation Platform Stack: Python • LangGraph • IBM WatsonX • FastAPI • IBM OpenPages • RabbitMQ • Docker • Kubernetes • Redis • MCP Protocol • Designed and built Bowmen OS — a production-grade multi-agent AI operating system that enables compliance teams to create and deploy custom GRC automation agents through natural language, with no coding required. • Engineered a sequential three-agent agentic pipeline — Planner Agent, FlowsAgent, and GapAgent — that collaboratively extracts user intent, constructs validated DAGs, and surfaces semantic gaps before code generation. • Built a DAG-to-LangGraph compiler that auto-generates agent scaffolding, tool bindings, memory configuration, and AI orchestration logic; includes a micro-orchestrator powered by IBM Granite for runtime natural language navigation. • Standardized all agent-to-platform communication using Model Context Protocol (MCP); built a FastAPI MCP tool server exposing OpenPages capabilities as callable tools for deployed AI agents. • Implemented Human-in-the-Loop (HITL) review layer using stateful orchestration persisted in Redis, enabling compliance officers to inspect, approve, or override AI agent actions before changes are committed — ensuring full regulatory traceability. • Leveraged RabbitMQ for event-driven async agent execution and Docker/Kubernetes for container orchestration, enabling enterprise-scale, non-blocking agentic workflow processing.

Project Engineer at Wipro, Dec 2022 - Jun 2024 · 1 yr 7 mos, Hyderabad, Telangana, India, At Wipro, I built AI solutions using multi-agent systems to automate vector indexing, streamline incident resolution, and enhance data access. I developed a fully automated indexing pipeline with Azure DevOps and Qdrant, and led the creation of a RAG-based Incident Resolution Copilot using LangChain agents. I also automated SharePoint integrations and designed modular components for cross-project consistency—delivering scalable and efficient AI workflows.

Artificial Intelligence Engineer at Wipro, Dec 2022 - Jun 2024 · 1 yr 7 mos, Medic Assistant — Agentic Clinical AI Chatbot (Kencor Health, USA) Stack: Python • LangChain • Qdrant • Flask • Advanced RAG • Guardrails AI • NEMO Guardrails • Amazon EC2 • Developed and deployed a production clinical AI agent for medical professionals to query structured and unstructured patient records in natural language using an agentic RAG pipeline with Router Agent, Retriever Grader, Answer Grader, and Hallucination Grader. • Achieved F1 score of 0.68 on HotPotQA benchmark; validated using RAGAS. Enforced AI response safety using Guardrails AI and NEMO Guardrails; integrated Conversation Summary Buffer Memory for multi-turn sessions. Smart Indexer — Automated Vector Indexing Framework Stack: Python • Qdrant • Azure DevOps • SharePoint • Jina AI • BGE-Large Embeddings • Rerankers • Architected a production-grade, fully automated document ingestion and vector indexing pipeline integrated with Azure DevOps and SharePoint; leveraged BGE-Large embeddings and Jina AI rerankers for hybrid search (vector + keyword). • Designed as a modular, reusable AI component library — enabling rapid deployment across multiple projects with minimal reconfiguration. SMART OPS — Generative AI Tool for IT Incident Resolution Stack: Python • LangChain • RAG • Azure AI Services • Azure GPT-3.5 • Flask • Built a GenAI-powered web application accelerating IT incident resolution via hybrid semantic search over past incidents and a LangChain AI agent for automated root cause analysis from live component logs. • Secured the API with Flask JWT and Active Directory authentication; built a PII masking module to protect sensitive data before it reaches any LLM.

Intern at Wipro, Feb 2022 - May 2022 · 4 mos, India, Spearheaded the design and implementation of a Hospital Management System, automating core hospital functions such as patient records, appointment scheduling, and billing processes. Utilized Oracle Applications and SQL/PLSQL for database management and Forms and Reports to build user-friendly interfaces, improving hospital operational efficiency. Gained in-depth expertise in database design and workflow optimization while mentoring team members and ensuring project delivery within timelines.

Education

Bachelor's degree, Electronics and Communications Engineering, Srm University

Bachelor of Engineering, Electrical, Electronics and Communications Engineering, SRM Institute of Science & technology