● Software Developer · Backend & Data · Toronto, ON

I take real problems and turn them into working software.

Computer Science graduate working in Python and SQL, with production data experience from a university HRIS. I also work with AI models directly, mostly prompt engineering and LLM tooling, and I've built four applications that way. Currently studying for the AZ-900 and moving toward backend and cloud work.

● Available for cloud & backend rolesToronto (UTC-5)

now: studying for the AZ-900

View work Résumé (PDF)

whoami

namePeter Yousefi rolecs grad · backend & data corepython · sql · git datasql · databricks learningazure (az-900) · flask · docker regiontoronto (utc-5) status● available

Selected work

Things I've built

Four applications, each starting from a problem someone actually had. I scoped the requirements, defined how they should behave, and iterated until the full flow ran.

4 projects · 2026

Continuity storyboard generator interface

Continuity — AI storyboard generator

If you generate a storyboard one scene at a time, your main character comes out looking like a different person in every frame. Continuity takes a story, a style guide, and a character sheet, uses GPT to split the story into scenes, then builds each image prompt from all three so the same character carries through the whole sequence. There's a toggle that regenerates the same scenes without the character sheet, side by side, so you can see what it's actually doing.

My part worked out the three-document approach and added the review step so you can edit every prompt before spending image credits. I also handled getting it live on Vercel, which is where the real problem showed up: image generation takes 15 to 40 seconds and Hobby serverless functions cut off at 10, so the image route runs as an Edge Function with a 60-second timeout that shows an actual error instead of hanging.

Next.js 14TypeScriptReactOpenAI APIVercel EdgeTailwind

Live app ↗GitHub Demo video

Lumina — AI campaign generator

Small business owners were losing hours on each campaign — writing copy, sourcing images, and tying promotions to local events. Lumina turns a business description into a ready-to-publish campaign in minutes, with every field owner-editable before it goes live.

My part scoped the problem and defined the behaviour — owner-editable output, and a fallback path so it keeps working when the API is down. Tested and iterated until the full flow ran.

PythonFlaskREST APIsIBM watsonx

GitHub Demo videoIBM AI Builders Challenge

Sharp — betting picks dashboard

Bettors were juggling odds screens, matchup research, and parlay math across separate tools. Sharp pulls live NFL odds, scores, and player props into one dashboard, with a bet-slip tracker, automatic P&L/ROI, and a responsible-gambling age gate on by default.

My part specified the dashboard, the bet-slip tracking, and the responsible-gambling age gate that is on by default.

SQLiteIBM watsonxReactNode/Express

GitHub

GeoShield — civic notification tool

Coordinating a prescribed burn meant manually working out which municipality a boundary touched, finding the mayor and fire chief's contacts, and drafting a notice by hand. GeoShield collapses that chain into a single run and drafts a calm, professional notification.

My part defined the notification chain and required the human-in-the-loop step that flags unverified contacts so a person reviews before anything sends.

.NETIBM watsonxLlama 3.3GIS

GitHubIBM AI Builders Challenge

What I work with

Stack

what I can actually talk through

Core

PythonSQLGit & GitHub

Data

SQL reporting & validationProduction HRISDatabricks

AI applied

Prompt engineeringLLM toolingIBM watsonxOpenAI API

Web & deploy

Next.jsTypeScriptReactVercel

Currently learning

Microsoft Azure (AZ-900)FlaskREST APIsBicep / TerraformGitHub ActionsDockerLinux & Bash

Working toward the AZ-900 certification, then hands-on Azure infrastructure deployment, infrastructure-as-code, and CI/CD.

Track record

Experience

Jul 2024 – Present

Avalogistics

Logistics Coordinator — Toronto, ON

Renegotiated 3PL carrier contracts, cutting overall freight costs by roughly 12% without sacrificing transit times.
Built a tracking routine to flag at-risk shipments earlier, giving sales and purchasing lead time to notify customers proactively.
Tightened commercial-invoice and customs review, reducing avoidable hold-ups at the border.

Sep 2025 – Dec 2025

Trent University

Grader — Oshawa, ON

Left specific comments on every assignment pointing to the exact concept behind each mistake, instead of just marking it wrong.
Ran one-on-one debugging sessions focused on teaching students to isolate a bug themselves rather than handing over the fix.

Apr 2022 – Sep 2022

Concordia University

HRIS Intern — Montreal, QC

Wrote SQL queries against the university's HRIS to build reports on employee and payroll data, joining across tables to assemble views the built-in reporting did not cover.
Monitored data replication between HR systems during an active ERP modernization, verifying records transferred intact and flagging discrepancies before they reached payroll.
Translated requests from HR staff and HRIS analysts into precise query criteria, confirming the output answered the underlying question rather than the initial phrasing.

Sep 2021 – Dec 2022

SCS & ECA, Concordia

VP of Social / Social Director — Montreal, QC

Rebuilt the event calendar around member feedback and delegated ownership across an 8-person committee — participation grew 40% and the society won Event of the Year twice out of 12 competing societies.
Was the on-the-ground point of contact for 200+ attendees and a rotating volunteer team, solving problems in real time.

Who I am

About

I'm a recent Computer Science graduate who likes taking a messy, real-world problem and turning it into something people can actually use — then making sure it holds up once it's running.

My strongest hands-on work has been with data: writing SQL against a live HRIS during an ERP migration, where a bad record reaches someone's paycheque. Alongside that I've taken four ideas from a blank page to working applications using AI builder tooling. The most recent one, Continuity, is live on Vercel, and getting it there meant fixing a serverless timeout problem rather than just shipping the happy path. Right now I'm studying for the AZ-900 and starting to build on Azure directly, so that the infrastructure side becomes something I've done rather than something I've read about.

Before this I coordinated international logistics and led a student society, which is probably why I care as much about the person using a system as I do about the system itself.

educationB.Sc. Computer Science, Trent University (2026)

locationToronto, ON

focusBackend / Data / Python

statusOpen to opportunities

Credentials

Certifications

in progress Microsoft Azure Fundamentals (AZ-900)

earned AWS & Databricks Cloud Integrations

earned Python 101 for Data Science

earned Data Workshop Level 2

Get in touch

Let's build something.

Open to backend and cloud roles: cloud support, junior cloud engineer, or backend-leaning positions. If you've got a real problem worth another set of hands on, I'll answer.

peter@petery.org

Send email LinkedIn GitHub