Shanmukha Jwalith Kristam

About

Built and deployed an AI agent to help patients navigate complex housing information by scraping and normalizing unstructured data across all 50 U.S. states, then layering a LangChain RAG system with MMR re-ranking to reduce hallucinations. Experienced in orchestrating multi-agent workflows (LangGraph/CrewAI) and production reliability practices (Pydantic-validated outputs, LLM-as-judge evals, tracing). Also delivered stakeholder-facing explainability via SHAP dashboards for a loan-approval predictive model at Welspot.

Experience

Education

Key Strengths

• Deployed production AI agent using distributed scraping + RAG across 50-state data sources
• Improved retrieval accuracy and reduced hallucinations via MMR re-ranking and better chunking
• Designed human-in-the-loop validation with async queues to avoid pipeline bottlenecks
• Builds reliable agent workflows with deterministic structured outputs using Pydantic
• Evaluation-driven development using LLM-as-judge, golden datasets, and metric tracking (accuracy/latency)
• Strong debugging/observability practice using tracing and span-based error analysis
• Pragmatic model selection balancing frontier models vs smaller fine-tuned models for cost/latency
• Bridges stakeholder trust gaps with explainability (SHAP dashboards) for high-stakes decisions