Ridhul Krishna P

AI Engineer with 2+ years of experience designing and deploying LLM-based and agentic AI systems in production. Specialized in retrieval-augmented generation (RAG), multi-agent orchestration with LangChain/LangGraph, and MCP-based tool-calling assistants, backed by a full-stack engineering foundation (.NET, Angular, Azure) that lets me ship AI features as reliable, monitored, end-to-end services rather than notebooks. Experienced in applying guardrails, structured outputs, and continuous-feedback loops to keep production LLM behavior safe and high quality.

AI / Software Engineer at Cloud Destinations (2024-07 – Present)

• Designed, trained, and deployed AI and deep learning pipelines using TensorFlow and Hugging Face for enterprise predictive and intelligent-automation use cases.
• Built an AI-powered virtual assistant integrated with MCP servers for self-directed task execution, knowledge retrieval, and natural language understanding, improving support automation efficiency.
• Productionized LLM applications using LangChain and LangGraph, implementing guardrails, validation layers, and controlled tool execution to ensure safe, high-quality outputs.
• Built retrieval-augmented generation (RAG) systems with vector search and semantic indexing to improve accuracy and contextual reasoning in AI assistants.
• Designed autonomous, multi-agent workflows for task planning, decision-making, and orchestration across enterprise systems.
• Analyzed model behavior, pipeline failures, and data issues to identify root causes and improve model performance and reliability; implemented monitoring pipelines to support retraining and iterative enhancement.
• Built scalable backend services (.NET, EF Core, Dapper) to support AI pipelines with secure data access and high-throughput inference, and integrated automated testing/deployment into Azure DevOps CI/CD.

Bachelor of Technology in Computer Science and Engineering (2020-06 – 2024-06)