Full-stack software engineer

Wasi Raza

Three years shipping data pipelines, APIs, and systems that ran in the real world, not just a demo. Bachelor's in Computer Science, currently teaching the next generation of developers to do the same.

Things I've shipped

Four systems, countless innovations and lessons.

NOVA desktop assistant screenshot
Fig. 01 — NOVA, activity + conversation view

Desktop assistantNOVA

Problem
Voice assistants either sound smart or actually do things — rarely both.
What it does
A voice-first desktop assistant built on Claude that reads native apps and the browser through Windows UI Automation and Playwright, keeps memory across sessions, and writes its own tools at runtime when it needs one it doesn't have.
Decision
Gave it real OS-level control instead of a sandboxed plugin API, so it can act inside apps that were never built with an assistant in mind.
agentic-aiclaude-apiwhisper
View NOVA repo on GitHub
CapsuleGuard_AI live computer-vision inspection feed
Fig. 02 — CapsuleGuard, live inspection feed

Computer visionCapsuleGuard_AI

Problem
Manual visual QC on a production line doesn't scale, and it doesn't stay consistent across a shift.
What it does
A computer-vision monitor that classifies items in real time and flags unsafe or defective ones instantly, built alongside coursework in generative AI engineering.
opencvpytorchreal-time-vision
View CapsuleGuard_AI repo on GitHub
AI-Incident-Analysis-Engine pipeline diagram
Fig. 03 — AI-Incident-Analysis-Engine, pipeline diagram

Data pipelineAI-Incident-Analysis-Engine

Problem
Incident reports pile up faster than anyone can read them for patterns.
What it does
Ingests raw incident reports, runs anomaly detection over them, and produces structured, model-assisted root-cause analysis instead of another dashboard nobody opens.
anomaly-detectionpythonincident-management
View AI-Incident-Analysis-Engine repo on GitHub
Event-Ingestion-Platform API response screenshot
Fig. 04 — Event-Ingestion-Platform, API response

Ingestion serviceEvent-Ingestion-Platform

Problem
Streaming event data breaks quietly — a dropped or duplicated write doesn't throw an error, it just costs you data.
What it does
A high-throughput ingestion service built on the same instincts as the production pricing pipeline I maintained at ZE PowerGroup: idempotent writes, and schema changes that don't take the system down.
javaspring-bootpostgresql
View Event-Ingestion-Platform repo on GitHub

// off the clock

Outside of shipping backends I'm usually in the middle of some other system I can't leave alone — right now that's 3D printing and electrical engineering, pulling apart how things are actually built instead of just how they're coded. Same instinct that makes me track down a flaky parser makes me want to understand why a circuit or a print failed, not just that it did.

What people say

From managers and teammates across teaching and engineering roles.

"Wasi is an exceptional developer and teammate who quickly adapted to challenging situations, demonstrating strong skills and a proactive learning attitude. His collaborative problem-solving approach not only accelerated solutions but also fostered mutual learning. Over time, he showed significant growth, taking on increasingly complex tasks independently. His professionalism, teamwork, and respect for others made a lasting impact, and he would be a valued addition to any team."Benjamin Tisserand — My supervisor at ZE PowerGroup, now Agile Developer II at SAP
"Wasi is a great team player who gets along well with everyone, which creates a positive work environment. He asks insightful questions that demonstrate his commitment to fully understanding projects and contributing effectively. Wasi's proactive approach and strong interpersonal skills make him a valuable asset to any team."Tony Huang — Java Developer, ZE PowerGroup
"Wasi is very helpful and courteous with instructors, students and CTL staff and does well to work with the many personality types and expectations in a multitude of scenarios and contexts that arise. His technical knowledge and ability to problem solve are an important asset to the success of the online synchronous sessions. Wasi is customer service oriented, a good team player and has a calm demeanor. Wasi has been a valued member of the support team for the MSN program."Janine Hirtz — Senior Educational Consultant, UBCO

Where I've worked

Four stops, in order.

Apr 2024 —
Jun 2025

Coding Instructor · Impact A&C

  • Taught programming fundamentals K–12: Python basics, app development, Scratch, and game modding.
  • Designed independent computer-vision and ML-based monitoring projects, alongside coursework toward the IBM Generative AI Engineering Professional Certificate.
Feb 2023 —
Apr 2024

Software Engineer · ZE PowerGroup

  • Maintained a large-scale production database ingesting real-time energy and commodities pricing data from third-party market vendors.
  • Diagnosed and resolved data-parsing failures by updating regex-based parsers as upstream vendors changed formatting.
  • Built an automated CI/CD test suite covering 95% of previously manual test cases.
javaspring bootsqloracle dbliquibasejenkins
Jan 2021 —
Apr 2021

Software Engineering Intern · Atomic47 Labs

  • Developed, tested, and documented Nest.js API endpoints for profile management and crypto transaction retrieval.
  • Redesigned a multi-step sign-in flow into a streamlined React portal, reducing the clicks required to log in.
nest.jsreactpostman
Sep 2019 —
Dec 2020

Full Stack Developer · University of British Columbia

  • Built a scientific research platform in React and TypeScript to handle data from thousands of Canadian clinics.
  • Delivered full SDLC for secure profile management and REST APIs, reducing corrupt entries via field-level validation.
reacttypescriptdocker

Hiring a full-stack engineer?

Open to backend- or frontend-leaning roles. Feel free to message me and let's set up a dialogue.

Fastest way to reach me. Hope to hear from you soon 😁