Hi, I'm Aman
I build software& AI systemsthat matter.
Software & AI engineer specializing in building production-grade AI applications, scalable backend systems, and polished web experiences.
By the numbers
About Me
Building at the intersection of software engineering and AI.
I'm Aman, a software & AI engineer with 4+ years of experience building systems that go from idea to production. I specialize in designing intelligent backends, fine-tuning and deploying Machine Learning models, and crafting user-friendly solutions.
I care deeply about code quality, system reliability, and building products that solve real problems. When I'm not writing code, I'm exploring new technologies, watching F1 or building my Minecraft world.
My Journey So Far:- Software Developer (Self-Initiated) — NovaQuazar Technologies (2025–Present)
- Data and Reporting Analyst (Academic Assistant) — Niagara College Toronto (2024–Present)
- Data and Systems Automation Engineer Intern — Nuance Communications (Microsoft) (2022–2023)
- Full Stack Developer Intern — Amity University (2020)
Let's work together→Interests & Focus Areas
“The best code is one that solves the right problem simply, reliably, and safely.”
Amandeep Bhalla
Software & AI Engineer
Toronto, ON
Technical Skills
Tools of the trade.
A curated set of technologies I use to build reliable, scalable, and intelligent systems.
Languages
AI / ML
Web & Backend
Infra & Tooling
Featured Work
Selected projects.
A handful of things I've built — from AI pipelines to developer tooling.
ManuFast - Small Language Model for Predictive Maintenance AI
A ~30M parameter Small Language Model built entirely from scratch in PyTorch, designed to help factories predict machine failures before they happen. It reads sensor data as plain English and delivers fault detection, remaining useful life estimation, and plain-English diagnostic summaries - all in a single forward pass, with no Hugging Face or pre-trained components used anywhere.
Project DataMind - AI Powered Data Analytics Multi-Agent System and Custom MCP
Most people drown in spreadsheets. DataMind flips that. Upload any CSV or Excel file and a pipeline of five AI agents instantly takes over, surfacing insights, detecting anomalies, and building a fully interactive analytics dashboard. Under the hood it's a custom multi-agent system built with LangGraph and a purpose-built MCP server, with natural language SQL querying, real-time web fact-checking, and a production-ready deployment.
Project DocBot - AI Based Document Q&A RAG System
This is an AI-powered document Q&A app where users upload a PDF and ask natural language questions about its content. Built on a production-grade RAG pipeline, implementing vector search, reranking, and LLM grounded generation. The project covers the full lifecycle from data ingestion to a deployed, secure, rate-limited web application across Railway and Vercel. Pipeline quality was validated using Ragas evals.
ProdPocket AI
A production AI platform for creators to produce SEO optimized content and images for better reach across various platforms including social media, blogs, and more.