RamKumar Manoharan

I have over 16 years of experience across software engineering, data science, machine learning, and AI, with recent hands-on specialization in Generative AI and agentic systems.

I design and develop Python-based AI agents and multi-agent workflows using LangChain, LangGraph, CrewAI, OpenAI APIs, and workflow automation platforms such as n8n. My experience includes building supervisor-agent architectures, defining agent states and routing logic, implementing tool and function calling, managing shared context, and designing human-in-the-loop approval workflows.

I have worked on agentic AI use cases involving research automation, intelligent response systems, cybersecurity monitoring, supply-chain operations, inventory and order coordination, education, and enterprise knowledge assistants. I have also developed RAG pipelines using vector databases, embeddings, document retrieval, prompt engineering, evaluation, and knowledge-graph-enhanced retrieval with Neo4j. I have practical knowledge of agent interoperability and context protocols, including Model Context Protocol (MCP) for tool integration and Agent-to-Agent (A2A) communication for multi-agent coordination. My work also covers guardrails, structured outputs, error handling, retries, observability, evaluation, and reducing hallucinations in LLM applications.

Senior Manager at TensorLearners, a consulting startup that helps organizations adopt and effectively use advanced machine learning and artificial intelligence technologies

Product Management and Solution Architecture - Generative A.I

Graph DB Integration with Gen A.I product

My role and responsibilities services include consulting clients on MLOps adoption, advanced machine learning approaches, AI technology architecture design, intelligent response system development, and corporate training. Worked on integrating LLM models for workflow using packages such as langchain, LLamaIndex