Applied AI,
engineered to ship.
We are a Canada-based applied-AI studio. We design and deliver production AI systems — computer-vision safety platforms, privacy-first document intelligence, and LLM knowledge assistants — from first architecture to live operation.
Waterloo-grad founders
AI & systems engineering
≈0.95
vision-vs-sensor accuracy
Canada · NL
delivered software exports
From first architecture to live operation.
Computer-Vision Safety (HSE)
Real-time monitoring on existing cameras — PPE detection, incident detection, and ergonomic risk scoring.
- Markerless motion analysis from ordinary phone or CCTV cameras
- Seven recognized ergonomic standards (REBA, RULA, NIOSH and more)
- On-prem or in-browser processing to keep sensitive video private
Privacy & Security AI
Document intelligence that finds and redacts sensitive data while keeping records usable.
- OCR + NLP detection of personal and sensitive information
- Human-in-the-loop review with editable, queryable results
- Compliance-aware workflows and secure object storage
LLM Knowledge Assistants
Retrieval-augmented assistants and organizational knowledge graphs over your own documents.
- RAG over technical docs, contracts, drawings, and email
- Domain-tuned language models with citations and traceability
- Custom engineering calculators and semantic search
Full-Stack Product & Advisory
End-to-end delivery and technical advisory — from architecture to live operation.
- System design, implementation, operation, and delivery
- Microservices, CI/CD, and edge-compute (WebGPU) architectures
- AI strategy and technical due-diligence
Production systems across three domains.
We keep client identities private. What follows is the shape of the work and the outcomes.
HSE
Ergonomic Risk Platform
A markerless, camera-only motion-capture tool for ergonomic assessment. Computes 3D joint angles and tracks multiple subjects, scoring posture risk in real time and producing actionable reports.
Outcome — ~0.95 correlation with sensor-based capture; privacy preserved via face/background blurring.
Smart HSE Site Monitoring
Real-time safety monitoring layered onto existing CCTV: PPE detection, unsafe-behavior detection, and incident detection (such as a person down) with instant alerts.
Outcome — Runs on cameras clients already have — no new hardware required.
Smart Kitchen Safety OS
An AI operating system for food-safety oversight that turns periodic inspections into persistent, 24/7 intelligence. It fuses computer vision — 3D digital-twin reconstruction, object detection, action recognition, and pose estimation — with IoT sensors to continuously monitor handwashing, cross-contamination, PPE, cold-chain temperature, and staff wellness, then surfaces everything in a single command center.
Outcome — Privacy-preserving by design — on-device skeletal tracking and anonymization send encrypted metadata, not raw video — deployable from a single home kitchen to large industrial operations.
AI Platforms
Construction Regulation Assistant
A microservices platform that digitizes national building regulations and delivers context-aware technical consultation via a retrieval-augmented LLM, a custom OCR pipeline, and interactive engineering calculators.
Outcome — Context-aware answers grounded in an organizational knowledge graph.
Automotive After-Sales Intelligence
An integrated LLM platform for a national automotive after-sales network. Four cooperating AI modules — a customer portal, a service-bay diagnostic assistant, a dealer workflow agent, and a multimodal technical knowledge base — standardize fault diagnosis, cut warranty costs from repeat repairs, and put instant, cited technical knowledge in every technician's hands.
Outcome — Lower warranty cost from fewer repeat repairs, higher customer satisfaction, and a dealer network empowered with instant, cited technical knowledge — on a continuously self-improving feedback loop.
IoT Smart-Metering Platform
A scalable IoT platform for real-time monitoring, leak detection, and automated control over low-power long-range (LoRaWAN) networks, built on a CQRS microservices architecture.
Outcome — Cut manual inspections and improved incident response times.
Idea to product, in four steps.
Discovery & need-finding
We start from a concrete need and root-cause the real challenge — whether that's search across massive documentation or the risk of leaking sensitive data.
Architecture & design
We design a technical architecture matched to the problem (knowledge graph, edge processing, secure storage) to guarantee accuracy and meet security requirements.
Prototype & pilot
We measure accuracy, correctness, and processing speed under real conditions, then fold operational feedback back into the product.
Delivery & operation
We ship the final product and deploy it into your infrastructure with CI/CD, then operate and support it.
Markerless, hardware-free
Accurate motion analysis (≈0.95 correlation vs. sensors) using only ordinary phone or CCTV cameras.
Privacy by architecture
WebGPU edge processing runs video in the user's browser — sensitive footage never leaves the client.
Standards depth
Simultaneous support for seven recognized ergonomic standards, where competitors support far fewer.
Own engines
In-house OCR, plate-reading, and ergonomic-scoring engines — and a domain knowledge base for engineering.
A small team that ships real AI.
Founded by University of Waterloo graduates, with engineers drawn from other top global universities. The team specializes in machine learning, computer vision, and large-scale software engineering.
- Peer-reviewed research on validation of markerless vision-based motion capture
- Export track record: software delivered to clients in Canada and the Netherlands
- International technical partnerships across Canada and Europe
Founder & CEO
University of Waterloo — Artificial Intelligence
Sets technical direction and leads applied-AI research across vision and language systems.
Co-founder & Chair
Top-tier CS program — Artificial Intelligence (MSc)
Leads engineering practice, delivery, and the company's MLOps and infrastructure standards.
AI Research Engineers
Graduates of leading global universities
Build the vision and language models that power the studio's products and client work.
Full-Stack & Platform Engineers
Experienced product and infrastructure builders
Turn models into reliable, secure products and operate them in client environments.
Tell us what you're building.
The first consultation is free. Use the assistant in the corner, or reach us directly.