skrypt.xyz/chaitanyahivlekarFremont, CA

AI Research Engineer

hivlekarchaitanya@gmail.com
Scroll

Education

  • Aug 2022 — May 2024

    University of Florida, Gainesville

    Master of Science · Computer and Information Science - AI Specialization
  • July 2018 — June 2022

    MIT-WPU, Pune

    Bachelor of Technology · Computer Science and Engineering

Skills

PythonRustJavaC++C#RTensorFlowPyTorchHuggingFace TransformersLangChainLangGraphLlamaIndexVLLMCLIPVITOpenCVYOLOBERTVector DBsFAISSMilvusIDEsCursorKiroVS codeIntelliJSQLMySQLPostgreNoSQLCassandraFastAPIStreamlitGitDocker

Experience

  1. July 2024 — Present

    Machine Learning Engineer

    Walmart · Sunnyvale, CA
    • Led design and development of a hybrid agentic architecture for Sam's Club merchandising using LangGraph, combining Router-Specialist delegation with intra node Planner-Tool-Critic loops. Enabled dynamic context injection and memory aware tool orchestration via MCP to diagnose product availability issues in natural language which reduces investigation time from hours to minutes and supports autonomous remediation with human-in-the-loop control.
    • Engineered a 24/7 autonomous support agent with webhook based integration (JIRA, ServiceNow, Confluence, Slack) and a hybrid RAG pipeline using semantic search (OpenAI + FAISS) for context grounded resolution which achieved 95% auto ticket resolution with retrieval enhanced reasoning and zero manual intervention.
    • Built and deployed an end-to-end anomaly detection system for 1.4M SKUs, saving $4M+ by flagging pricing and dimensional errors. Integrated a vision pipeline (YOLOv8 + reference scaling) to estimate product dimensions from images, and used LLM-based parsing to extract dimensions from specs sheets. This boosted model accuracy by 38% and cut manual QC by 95%.
    • Developed a vision + LLM pipeline to extract nutrition facts from product labels at Walmart and Sam's Club using OCR (Tesseract) and GPT-5 for SNAP eligibility classification.
    • Built a dynamic Model Context Protocol (MCP) framework spanning multiple resource tiers enabling self discoverable services and orchestration for agent based automation at scale.
  2. January 2023 — December 2023

    Research and Teaching Assistant (ML, Bioinformatics, and Database)

    University of Florida
    • Processed and analyzed 10+ TB of human genome data to identify genetic patterns associated with ALS.
    • Designed a custom statistical framework for novel repeat expansion detection currently under lab validation for clinical research.
  3. June 2021 — October 2021

    Machine Learning Intern

    Whizkey
    • Engineered a signal disaggregation pipeline using NILM techniques and time series ML models, paired with a real time Streamlit dashboard which reduced monitoring overhead and enabled predictive maintenance, extending appliance lifespan in 86% of deployments.
  4. July 2021 — September 2021

    AI and Data Science Intern

    Hornbill Labs
    • Designed an ensemble time series forecasting pipeline with a Plotly analytics dashboard to optimize retail inventory flow-achieved 98% forecast accuracy and delivered comparative model insights for supply chain planning.

Projects

  • 01

    AI Fitness Vision Coach

    OpenCV, YOLO, LangChain Agent, HuggingFace Transformers

    • Built an agentic AI system that uses pose estimation and real time computer vision to evaluate exercise form, count reps, and adapt workouts dynamically.
  • 02

    ScholarAI

    LangGraph, LlamaIndex, FAISS, HuggingFace Transformers, Streamlit

    • Developed an agentic AI system integrating a self-adaptive tutor and autonomous research assistant.

Contact