Industrial Anomaly Detection Pipeline
An end-to-end system using autoencoders to flag product defects in real-time. Features configurable models, detailed performance logging, and visual heatmaps for quick diagnosis.
Hello, I'm Raj Chovatiya, an AI Engineer.
I build scalable AI systems & seamless digital experiences.
Specializing in Artificial Intelligence, Software Development, Vibe Coding and turning complex data into intelligent, user-centric solutions.
- Python
- JavaScript
- TensorFlow
- PyTorch
- React.js
- FastAPI
- Docker
- Vector DB
- LangChain
- RAG
- Python
- JavaScript
- TensorFlow
- PyTorch
- React.js
- FastAPI
- Docker
- Vector DB
- LangChain
- RAG
About MeFull-Stack Engineer with an AI Focus
I am a results-driven Software Engineer with over a year of experience building production-grade applications that serve thousands of daily users. My expertise lies at the intersection of robust backend architecture and intelligent systems.
From designing distributed systems for financial data processing to developing high-accuracy deep learning models, I am passionate about leveraging technology to solve complex problems. I thrive in collaborative environments and am committed to continuous learning and applying cutting-edge AI techniques to create impactful products.
Featured WorkCurated Case Studies
NeuralBlack: Medical Diagnostics AI
A high-accuracy CNN model (99% precision) using ResNet50 and Transfer Learning for medical image classification, deployed with a responsive web interface for real-time diagnostics.
YouTube Content Analysis Platform
An interactive platform leveraging custom LLMs and RAG for deep analysis of large-scale video content, with an optimized, real-time query pipeline for multimedia datasets.
AI-Powered Stock Analysis Tool
A Flask-based tool that aggregates data from multiple financial APIs to generate comprehensive, AI-driven investment reports for over 5000 tickers, empowering investors with actionable insights.
Video Detection Summary Engine
Automated system processing video streams with YOLOv5 to generate per-frame JSON outputs. It summarizes findings, identifies key frames with maximum diversity, and visualizes object frequency.
Model Comparison & Deployment
A framework for benchmarking object detection models (e.g., YOLOv5n vs. YOLOv5s) on inference time and accuracy. Includes a fully containerized Docker deployment for portable, scalable inference.
Professional Experience
AI Engineer
Tankhwa Patra | Oct 2025 – Present
Joined the core AI team to architect and implement advanced machine learning systems, contributing to the development of next-generation financial technology solutions.
AI Engineer
Grabsolve Infotech | Feb 2025 – Oct 2025
Built and deployed production-grade AI applications using Python and GCP, reducing operational inefficiencies by 25%. Developed automated forecasting systems with 95% accuracy and optimized ML model inference performance by 40% through advanced cloud-native architecture.
Machine Learning Engineer
Miracles Fintech Private Limited | May 2024 – Jan 2025
Led development of production-scale applications using Python & FastAPI, handling over 10,000 daily requests and architecting distributed systems for financial document processing.
Software Developer Intern
Drodify Proprietary Firm | Jan 2024 – May 2024
Contributed to multiple product cycles, developing Python-based solutions and implementing containerized applications with Docker and PostgreSQL.
Education
Institute of Technology, Nirma University
Master of Computer Applications (MCA)
Let's build something great together.
I'm open to new collaborations and challenging projects. If you have an idea you'd like to discuss, feel free to get in touch.