Vishwa J
AI Engineer
Overview
Social Links
About
I build production-grade AI systems that solve real business problems and ship reliably. My work spans machine learning, deep learning, LLMs, and AI agents, focusing on end-to-end systems—from data pipelines and model training to scalable deployment and monitoring.
I engineer solutions with a production-first mindset, optimizing for constraints, cost efficiency, and speed of delivery. This includes designing architectures that reduce inference costs, scale to real workloads, and deliver measurable impact.
GitHub Contributions
Stack
Experience
HYPERVERGE
Current Employer- Developed and released production features across internal AI services, including Generic Forgery Check, Name Match, and Centralized KYC systems, while contributing to model evaluation, performance benchmarking, and continuous monitoring, improving reliability, extensibility, and accuracy of identity verification workflows.
- Built and deployed an Application Form Validation API using AWS Rekognition for document OCR and face comparison, integrating fuzzy matching for data-to-document verification and containerizing the service with Docker for serverless deployment on AWS Lambda.
- Engineered a Slack bot for support engineering query tracking used by 300+ members, streamlining incident triage and improving SLA compliance for issue resolution.
- AWS Rekognition
- AWS Lambda
- Docker
- OCR
- Machine Learning
- Computer Vision
- Python
- Model Evaluation
- API Development
FORGE INNOVATION & VENTURES
IQUBE – INNOVATION CENTER
Education
Projects(6)
An advanced tool for generating synthetic Aadhar cards (Indian government ID) using both computer vision techniques and AI-based generative models. This project implements two distinct approaches to ID generation:
- CV-based Generation: Uses OpenCV and EasyOCR to modify existing Aadhar card templates with new information.
- AI-based Generation: Leverages Stable Diffusion 3.5 Large and Flux.1 Dev to generate completely synthetic IDs.
Key Features:
- Natural Language Processing to extract details from free-form text
- Dual Generation Methods (CV-based or AI-based)
- Customizable blurring effects
- Simple Gradio web UI
Disclaimer: This project is for research and educational purposes only.
- Computer Vision
- OpenCV
- EasyOCR
- Stable Diffusion 3.5
- Flux.1 Dev
- Python
- Gradio
