About me
Software engineer with 5+ years of experience building scalable backend systems and shipping machine learning solutions to production. I currently lead AI initiatives at EnerZam across energy and CMMS platforms, focusing on ML pipelines, model deployment, and AI platforms that operate reliably at scale. Click through the questions to learn more about my background, role, and interests.
I'm a software engineer with over 5 years of experience building scalable backend systems and delivering machine learning solutions in production. My work combines engineering fundamentals with applied AI, focusing on ML pipelines, model deployment, and operational reliability.
Work Experience
Full-Stack Software Engineer
EnerZam Inc.
Toronto, Canada
- Led AI feature development across energy and CMMS platforms, from data prep to model integration and safe rollout.
- Built Python data pipelines to ingest, clean, and validate energy data from Kafka into PostgreSQL for ML training.
Graduate Research Assistant
Concordia University, CENPARMI Lab
Montreal, Canada
- Collaborated with a research team to tackle challenges in machine learning and deep learning.
- Took a leading role in developing deep learning-based research projects from concept to implementation.
Web Application Developer
0&1 Information Technology Inc.
Tehran, Iran
- Built core backend features for a web-based management system using C#, ASP.NET MVC, and SQL Server.
- Focused on data modeling and writing efficient T-SQL queries to support system performance.
Featured Projects
A full-stack operational monitoring platform that aggregates system metrics and events in real time, giving teams centralized visibility into system health through a resilient, API-driven dashboard.
- Time-window filtering and data segmentation
- Role-based access control with conditional UI rendering
- Graceful degradation under partial failures or delayed data
- Background jobs pre-compute metrics for low-latency queries
My Tech Stack
Competing at Hackathons
NVIDIA Spark Hack Series
Built Parcel-Brief, an on-device AI co-developer for Toronto real estate analysis. Clicking a parcel on a map triggers a LangGraph pipeline that streams a full development brief including zoning, 3D massing, financial pro-forma, and city council vote prediction, all running locally on a single NVIDIA GB10 GPU.
- Ran NVIDIA Nemotron-Super-120B entirely on-device via vLLM on a single GB10 GPU, no cloud API calls
- Orchestrated 6 parallel and sequential LangGraph agents covering zoning, massing, pro-forma, and approval prediction
The Team
Amin Shamshiri
Full-Stack Engineer- Built the Next.js map interface and brief viewer
- Wired the parcel map to the streaming backend
- Owned the frontend pipeline from click to brief
- Implemented the Plotly 3D massing visualization
From the Hackathon Floor
A look at the team in action at the Google Cloud & NVIDIA Spark Hackathons
Volunteer Contributions
TEDxLakeheadU
I built the entire TEDxLakeheadU website from scratch using Next.js, React, TypeScript, and Tailwind CSS, since this was a brand-new TEDx chapter with no existing site to build on. My work covered the homepage, the speaker lineup, the event schedule, and a countdown timer to opening night, all designed to look clean and load fast on both desktop and mobile. The site was the first thing attendees saw, so getting that first impression right mattered most.
