KL // Operator file 00 // Uplink Ottawa --:--:--Sync 000%
Gatineau Ottawa River Ottawa // Home base Rideau Canal

KyleLeBlanc

Full stack software engineer // Ottawa, Canada

I build scalable web applications end to end: React and TypeScript up front, Node.js, Prisma and PostgreSQL behind, with APIs and architecture designed to hold under load.

Lat45.4215 NLon75.6972 WElev071 mScanLive, move to survey
Chapter 01 // Loadout

Tools of the trade

Three specialisations, the kit behind them, and the AI stack I work with in 2026.

Slot A1Specialisation

Scalable full stack

High-performance applications with front and back end built as one system, tested and optimised.

  • React
  • TypeScript
  • Redux
  • Node.js
  • Prisma
Slot A2Specialisation

API development

Secure, scalable RESTful APIs designed for clean data exchange and easy deployment.

  • Node.js
  • Express.js
  • Prisma
  • REST
Slot A3Specialisation

System architecture

Architectures that balance scalability, performance and maintainability.

  • Microservices
  • Docker
  • Heroku
  • AWS
Slot B1Kit

Languages

  • JavaScript
  • TypeScript
  • Python
  • Java
  • C
  • SQL
Slot B2Kit

Frameworks

  • Next.js
  • Spring Boot
  • Django
  • Flutter
  • Bootstrap
Slot B3Kit

Data and testing

  • PostgreSQL
  • Redis
  • Bull queue
  • MongoDB
  • Jest
  • Enzyme
Slot C1AI // 2026

LLMs, local and cloud

I run models on my own hardware and in the cloud, and I know when each is the right call for privacy, cost, speed and quality.

  • Local models
  • Cloud models
  • Model selection
  • Cost and latency
Slot C2AI // 2026

Agentic workflows

Agents that plan, use tools and finish multi-step work on their own, built with OpenClaw and similar frameworks.

  • OpenClaw
  • Tool use
  • Multi-step agents
  • Automation
Slot C3AI // In training

Training and fine-tuning

Currently learning on my own servers: working with real data, training custom models from scratch and fine-tuning existing ones.

  • Self-hosted servers
  • Real datasets
  • Custom models
  • Fine-tuning
Chapter 02 // Mission log

The route so far

One posting, three ranks. Each one raised the difficulty.

  1. OriginTraining

    Self-taught

    No formal degree // Learned by building

    Where it started. I taught myself to code with whatever I could find: Harvard's free CS50 course, Udemy courses, YouTube, and a long run of my own projects, building and breaking things until they worked.

  2. Rank 01From Sep 2022

    Junior Software Engineer

    Habit App Inc. // Trace // Remote, Canada

    Built the React interface for Trace, an all-in-one sneaker consignment platform: reusable components, sorting and filtering. Added REST endpoints in Node.js and PostgreSQL, wrote front-end unit tests with Jest and Enzyme, and cut Shopify API errors by 40% with rate limiting.

  3. Rank 02Field promotion

    Temporary Project Lead

    Habit App Inc. // Trace

    Took the lead and went after performance. Moved heavy work onto a distributed Redis and Bull queue, added Node.js clustering, tuned the database, migrated the codebase from Sequelize to Prisma, mentored a new developer and documented the whole system.

  4. Rank 03To Jan 2024

    Full Stack Software Engineer

    Habit App Inc. // Trace

    Resolved long-standing issues, rewrote weak code, and shipped features from user and client feedback across React, TypeScript, Redux, Node.js and Prisma. Kept mentoring and reviewing new hires.

  5. SoloIndependent

    Founder Newly launched

    LeBlanc Software Solutions

    Helping businesses bring AI into their companies. That includes building custom fine-tuned models on a company's own data and running them locally, so sensitive information never has to leave the building and there is no legal exposure from handing it to a cloud provider. I make those decisions easy.

  6. BuildProduct

    DiscoverTonight.city

    Nightlife discovery platform

    A platform where people share what a venue is really like tonight: the prices, the cover charge, the vibe and how long the line is, so everyone else knows before they go.

    DiscoverTonight.city
Chapter 03 // Confirmed results

Numbers from the field

What my work changed for the business on Trace, in production.

99%

Faster servers

Heavy jobs were slowing the whole platform down. I moved them into background queues so the app stayed fast while the work ran, improving server performance by 99%.

50%+

Faster page loads

Pages were slow because the app kept asking the database for the same data. I cut the wasted requests and page load times dropped by more than half.

40%

Fewer Shopify sync errors

Syncs with Shopify were failing when the app sent requests too quickly. I paced the requests and fixed the sync logic, cutting errors by 40%.

Chapter 04 // Operations

Four operations on file

Client work, company work, and the projects that built the skills. Keep scrolling to flip through the file.

CompleteOp 001Sep 2022 to Jan 2024

Trace by Habit App Inc.

An all-in-one platform for sneaker consignment, built by a remote team across Canada. I worked across the whole stack, from the React interface to the queueing and database layers, and led the project for a stretch.

  • React, Redux
  • TypeScript
  • Node.js
  • PostgreSQL, Redis
  • Prisma
trace / dashboardScreens
Trace store dashboard showing sales, payouts and inventory valueTrace inventory search with filters and a list of consigned sneakers
ActiveOp 002Product

DiscoverTonight.city

A nightlife discovery platform. People share what a venue is really like tonight: the prices, the cover charge, the vibe and the line, so everyone else knows before they go.

DiscoverTonight.city
  • Prices
  • Cover charge
  • The vibe
  • Line length
discovertonight.cityScreens
DiscoverTonight.city home page: Know before you go, with live wait, cover and price reports
CompleteOp 003Redesign

Portfolio website redesign

The previous version of this site, rebuilt on Next.js to improve performance, security and maintainability.

  • Next.js
  • JavaScript
  • HTML, CSS
kyleleblanc.ca // 2024Preview
WELCOME
The Professional PortfolioKyle Jorden Rene LeBlanc
Scalable full stack
API development
System architecture
TrainingOp 004Foundations

Harvard CS50

The course that laid the foundation: problem solving with and without code, across domains from cryptography to finance to forensics.

  • C, Python, SQL
  • JavaScript
  • Django, React
pset1 / mario.cPreview
 1 #include <stdio.h>
 2
 3 int main(void)
 4 {
 5     int height = 4;
 6     for (int row = 1; row <= height; row++)
 7     {
 8         for (int s = 0; s < height - row; s++)
 9             printf(" ");
10         for (int b = 0; b < row; b++)
11             printf("#");
12         printf("\n");
13     }
14 }
$ make mario && ./mario
   #
  ##
 ###
####
$ 
Chapter 05 // Codec

Incoming transmissions

Three letters of recommendation, on an open frequency. Tune to hear each one.

Frequency140.85PTT

Chapter 06 // Open channel

Next mission: yours

Briefing complete. If you have a question or a project to discuss, send the coordinates.

kyle.leblanc88@gmail.com
Company: LeBlancSoftwareSolutions.com