Chandramouli Prabhakaran

Chandramouli Prabhakaran

Computer Vision Engineer

Followers of Chandramouli Prabhakaran629 followers
location of Chandramouli PrabhakaranNovi, Michigan, United States

Connect with Chandramouli Prabhakaran to Send Message

Connect

Connect with Chandramouli Prabhakaran to Send Message

Connect
  • Timeline

  • About me

    AI Engineer | Expertise in Prompt Engineering, RAG pipelines and LLM fine tuning | Azure cloud | Interest in High Performance Computing Systems

  • Education

    • Kent State University

      2023 - 2024
      Master's degree Artificial Intelligence
    • BSA Crescent Institute of Science and Technology

      2019 - 2023
      Bachelor of Technology - BTech Computer science and engineering
  • Experience

    • Chargeback Gurus, an Inc. 5000 company

      Jul 2022 - Dec 2022
      Computer Vision Engineer

      - Developed an algorithm on pytorch to produce scores on quality and legitimacy of evidence used in dispute resolution, leading to a 25% improvement in chargeback recovery rates. - Streamlined the evidence quality evaluation processes to reduce the dispute resolution team from 10 members to 3. Cutting labor costs by 70% while increasing operational efficiency. - Successfully met all KPIs set during the timeframe, demonstrating measurable improvement on productivity and delivering significant cost savings. Show less

    • LTIMindtree

      Jan 2023 - Jun 2023
      Computer Vision Engineer

      Contributed to Research and Development within a team-oriented setting, contributing actively to daily scrum meetings to innovate solutions for unique business use cases on geographical and spatial data.Led the implementation of customized ML models (YOLO, FasterRCNN, SSD) to monitor solar panel construction located in Saudi Arabia, leveraging ArcGIS API and Model Builders.Carried out the development of image provessing and segmentation models (SAM, Unet, MaskRCNN) with 79 % accuracy purpose-built for the detection and classification of agricultural fields (satellite image) for an IBM project. Show less

    • Smart Communities and IoT (SCI) Lab

      Aug 2024 - Jan 2025
      Research Assistant

      - Collaborated with Gastroenterology doctor Dr. John Park, MD to develop a LLM designed to run on Nvidia Jetson Nano Orin for QA based diagnosis of gastrointestinal and liver diseases. - Developed and deployed LLMs on edge devices, optimizing Llama 3.2's performance (benchmarked against GPT-4o) through a RAG pipeline using domain adaptation techniques that simulates human reasoning via query generation and knowledge graphs. Ongoing evaluation, with plans for publication in ACM journals.

  • Licenses & Certifications