Registered Canadian business
DIXKOX Inc. is a human-centred technology and intelligent systems company that designs and builds software, cloud, and AI solutions, supported by rigorous quality assurance, testing, and security practices to ensure reliability, performance, and trust.
Screenshots from the founder's own projects: real tools, real runs.
The founder at work and presenting project results. Click the arrows to look through.
Five areas of engineering, each with the tools used in the published projects below.
Web applications built end to end: the front end, back-end services, APIs and databases.
CI/CD pipelines, deployment automation and cloud infrastructure, so a change is built, tested and released the same way every time.
AI and machine learning features inside real products, with answers tied to their source and a refusal when the source does not support one.
Test strategy, automation, performance and usability testing for web, mobile and desktop software.
Security practice built into how software is delivered: threat models, web vulnerability testing and risk checks before release.
Every project follows one line, and it starts with the people who will use the system.
What DIXKOX uses to build, test, secure and deliver software.
Projects published on the DIXKOX GitHub, built by the founder during a Master of Science in Software Engineering at Quantic.
Ask questions of a policy PDF. Answers come from the document, and a question it does not support is rejected.
Went from 15 to 20 correct on a 20-question evaluation set after the retrieval fixes. 35 regression tests pass.
Python, FastAPI, Next.js, TypeScript, Vercel, RenderView on GitHub →Answers questions from 16 synthetic company policies and shows the source section for every answer.
On 50 questions: 90% correct, 90% citation accuracy, median response about 720 ms.
Python, FastAPI, TF-IDF, scikit-learnView on GitHub →A web app that gives a first check on whether an uploaded Windows executable looks like malware.
Seven model types compared with stratified 10-fold cross-validation on about 50,000 records. Tests gate every deployment.
Python, Flask, scikit-learn, PyTorch, YARA, GitHub ActionsView on GitHub →GitHub Actions pipelines in four repositories: tests, builds, and deployments to Render and GitHub Pages, plus Docker image builds.
Each pipeline runs on every push, so a failing test or build shows straight away.
GitHub Actions, Docker, GitHub Pages, RenderView on GitHub →
Founder and Full Stack Developer
Started in health, safety, security and environment work, then quality management systems, and learned to ask what can go wrong before it does. DIXKOX applies the same habit to software, cloud and AI.
Holds an Ontario College Graduate Certificate in Software and Information Systems Testing from Fanshawe College (December 2025) and has completed a Master of Science in Software Engineering at Quantic (October 2026).
Software, AI, cloud and cybersecurity work and partnerships are welcome. Write to:
info@DIXKOX.com