Jackiv Garg

Agentic AI Developer and Computer Science undergraduate with experience improving production LLM pipelines and building voice AI, agentic workflows, and deployable ML systems. Skilled in Python, LangGraph, LangChain, RAG, FastAPI, and MLOps, with a focus on reliable orchestration, evaluation, and measurable system performance.

Agentic AI Developer at Munshot (2026-04 – Present)

• Extended production agentic extraction pipelines with requirement-aware API and prompt routing, output validation, and edge case handling, reducing extraction errors by 18%.
• Built an end-to-end KPI extraction pipeline to extract and validate KPIs from PDF reports, achieving 92% extraction accuracy.
• Optimized multi-step LLM workflows and query-routing endpoints, reducing end-to-end latency by 25% and improving throughput by 30%.
• Built dashboards for pipeline runs, extraction accuracy, and failure patterns across 500+ daily runs, cutting mean time-to-detect regressions by 40%.
• Built a personalized user-memory extraction pipeline for downstream agent reasoning, improving response relevance and personalization accuracy by 15%.

B.Tech in Computer Science and Engineering – Bennett University (2024-08 – 2028-05)