Hi, I'm Ashay Kargaonkar

Software Engineer with 3+ years of experience building secure, scalable backend systems using C#, .NET 8, and ASP.NET Core. Specialized in cloud-native development on AWS (Lambda, S3, EC2, App Runner, Cognito, DynamoDB) with deep expertise in RESTful API design, OAuth 2.0 and JWT authentication, asynchronous processing, and CI/CD pipelines using GitHub Actions and Docker. Experienced in both SQL and NoSQL database integration, and frontend development using React, Next.js, and TypeScript. Contributed to production-grade applications serving over 1.5 million daily users, with a strong focus on performance, reliability, and clean maintainable code.

Also actively building in the AI/ML space with a specialisation in Computer Vision. Designed and deployed end-to-end CV pipelines using YOLO, TrOCR, ONNX, and OpenCV for real-world image processing and handwriting recognition. Experienced in dataset preparation and annotation using Roboflow, model training with PyTorch, and deploying inference services on AWS EC2 with GPU support. Leverage Claude as an AI coding assistant to accelerate development workflows. Passionate about bridging production backend engineering with applied machine learning to build intelligent, scalable systems.

Backend Engineer ResumeComputer Vision AI Engineer Resume

Work Experience

Software Engineer

Tata Consultancy Services

March 2022 – June 2025

Chicago, IL

Client: United Airlines

  • Migrated ~21 million records from SQL to AWS DynamoDB with zero data loss, improving scalability and performance.
  • Developed and optimized RESTful APIs using C# and .NET, integrated with AWS Lambda, Parameter Store, and DynamoDB.
  • Built frontend components with React and TypeScript, implementing dynamic routing and server-side rendering via Next.js.
  • Containerized applications using Docker and deployed to Kubernetes (EKS), streamlining CI/CD workflows.
  • Collaborated with DevOps to automate deployments using Harness pipelines and ensured secure coding with Veracode scans.
  • Debugged and tested full-stack features in coordination with QA, enhancing code reliability and user experience.
  • Used Postman for API testing, GitHub for version control, and AWS CloudWatch for system observability.

Tools & Technologies

.NETC#MVCNext.jsAWSDynamoDBLambdaS3PostmanDockerGitHubHarnessSwaggerCI/CDAgileRESTful APIsUnit TestingSystem DesignDatabase DesignCoordination with QA

Data Analyst Intern

Instahub LLC

January 2021 – March 2021

Philadelphia, PA (Remote)

  • Analyzed sensor data from AWS RDS (MySQL) to identify patterns in room usage and device activity.
  • Built predictive models using R and Python to forecast energy savings based on room occupancy.
  • Conducted statistical analysis to uncover trends, correlations, and support data-driven decision-making.
  • Designed interactive dashboards in Tableau to visualize sensor insights for business stakeholders.
  • Utilized SQL Workbench and Jupyter Notebook for data extraction, transformation, and model evaluation.

Tools & Technologies

AWS RDSMySQLPythonRmatplotlibTableauJupyter NotebookData AnalysisPredictive ModelingStatistical AnalysisEnergy OptimizationSensor Data ProcessingData VisualizationDashboard DesignTrend Analysis

My Projects

A showcase of my recent work

ThinkMoves

ThinkMoves

April 2025 - Present

Visit thinkmoves.com

Overview

ThinkMoves is an AI-powered platform that converts handwritten chess scoresheets into accurate, digital PGNs, making post-game analysis fast, reliable, and shareable. The platform automates the full pipeline — from reading handwritten move sequences to validating them against chess rules and reconstructing the complete game. Users can review and edit extracted moves where needed, eliminating manual data entry for players, coaches, and tournament organizers. Currently supporting the ThinkMoves scoresheet format, it is designed to scale and support more sheet formats for the broader chess community.


The platform is built on a multi-stage Computer Vision pipeline combining Image Classification, Object Detection, and Handwriting Recognition. It uses a custom-trained YOLO model for detecting and segmenting scoresheet regions, and a TrOCR model converted to ONNX format for fast, optimised handwriting recognition inference. Models are deployed on a GPU-based AWS EC2 instance powered by NVIDIA CUDA for low-latency processing. The backend is built with .NET 8 and ASP.NET Core, with an asynchronous event-driven pipeline on AWS Lambda and S3, and DynamoDB for structured data storage.

.NET 8ASP.NET CoreYOLOTrOCRONNXOpenCVNVIDIA CUDAAWS EC2AWS LambdaAWS S3DynamoDBCognitoReactNext.jsTypeScriptChess.jsVercel
AshayTalks

AshayTalks

Upgrade Coming Next

Overview

AshayTalks is a personal AI chatbot I built to give recruiters and collaborators a faster, more intuitive way to explore my professional background — no resume digging, no LinkedIn scrolling. Powered by Claude (Anthropic), it delivers accurate, context-aware answers about my work experience, technical skills, projects, and career journey. AshayTalks is a living reflection of my growth as a Software Engineer and AI/ML Engineer, continuously updated as I take on new projects and expand my skill set.

ReactNext.jsTypeScriptJSONGPT-4o-miniVercelGitHubVibe CodingAI Chatbot

Skills

Machine Learning & AI

TensorFlowPyTorchYOLOv8ResNetTrOCROpenCVRoboflowComputer VisionOCRMLOps

Backend & Cloud

C#.NET CoreASP.NET CoreREST APIsAWS S3AWS CognitoAWS App RunnerAWS LambdaDynamoDB

Frontend & Programming

PythonSQLTypeScriptJavaScriptReactNext.jsTailwind CSS

DevOps & Data Visualization

DockerGitHub ActionsTableauMatplotlibSeabornggplot2Power BI

Certifications

AWS

AWS Cloud Practitioner

2023

Udemy

View Certificate
AWS

AWS AI Practitioner

2025

Udemy

View Certificate
ML

Machine Learning A-Z

2023

Udemy

View Certificate
DL

Deep Learning A-Z

2023

Udemy

View Certificate
AWS

AWS Machine Learning Specialty

Ongoing

Udemy

Link to Course

Upcoming Projects

ThinkMoves

In Progress

A powerful tool to analyze and convert chess scoresheets into digital PGNs. Select access will be provided soon. Development is ongoing.

AshayTalks Upgrade

Coming Next

A new version of AshayTalks where users can talk directly to the chatbot. Personal interaction features are being integrated.