DevPilot AI
A full-stack SaaS web application for software developers. AI-powered assistant that automates code reviews, documentation, bug detection, and test case generation using LLMs.
Nihar Bhatt — AI/ML Engineer
Computer Engineering undergraduate at MIT Academy of Engineering with hands-on experience in Machine Learning and Computer Vision. Built real-time object detection systems using YOLO and developed data-driven automation tools. Currently open to opportunities.
┌───────────────────────┐ │ ███╗ ██╗██████╗ │ │ ████╗ ██║██╔══██╗ │ │ ██╔██╗ ██║██████╔╝ │ │ ██║╚██╗██║██╔══██╗ │ │ ██║ ╚████║██████╔╝ │ │ ╚═╝ ╚═══╝╚═════╝ │ │ │ │ > models: training │ │ > status: learning │ └───────────────────────┘┌─────────────────────────────────────┐ │ │ │ ███╗ ██╗██████╗ │ │ ████╗ ██║██╔══██╗ │ │ ██╔██╗ ██║██████╔╝ │ │ ██║╚██╗██║██╔══██╗ │ │ ██║ ╚████║██████╔╝ │ │ ╚═╝ ╚═══╝╚═════╝ │ │ AI/ML Engineer │ │ │ │ > models trained: active │ │ > status: open to work │ │ > focus: deep learning │ │ │ └─────────────────────────────────────┘
Artifacts
A full-stack SaaS web application for software developers. AI-powered assistant that automates code reviews, documentation, bug detection, and test case generation using LLMs.
Real-time weapon detection system using YOLO for surveillance applications. Trained and fine-tuned on custom dataset with OpenCV for real-time video processing and object localization.
App & web-based software for productivity and safety management in coal mines. Features attendance tracking, task management, incident reporting, and real-time monitoring dashboards.
Collection of responsive web interfaces and interactive UI components developed during SkillsCraft internship. Optimized for performance and user experience.
Field Notes
Brief observations, technical findings, and thoughts from the workbench.
Learnings from compiling the kernel, configuring BusyBox, and creating bootable ISOs with Syslinux.
Implementing Model Context Protocol for seamless AI model interactions with vector databases in RAG apps.
Exploring the new features in Next.js 16 and migrating to Tailwind CSS v4's new configuration system.
Running Llama2 locally and building a personal chatbot API for natural language tasks.
Work in Progress
Active experiments and prototypes. Things that are being built, broken, and rebuilt.
Full-stack SaaS for developers - AI-powered code reviews, documentation, bug detection, and test generation using LLMs
Real-time weapon detection system using YOLO for surveillance with custom dataset training
App & web-based software for productivity and safety management in coal mines