Computer vision Β· Spatial awareness
Ona
Object Navigation Assistant that turns camera input into object, depth, face, OCR, and scene state for real-time spatial awareness.
- Object detection
- Depth
- Face recognition
- OCR
Selected work
I work across computer vision, applied machine learning, backend systems, and product engineering. These projects are the clearest expression of how I think: start with messy real-world input, make it legible, then build something dependable around it.
Computer vision Β· Spatial awareness
Object Navigation Assistant that turns camera input into object, depth, face, OCR, and scene state for real-time spatial awareness.
Local-first AI Β· Vision and voice
A local-first visual and conversational assistant combining on-device vision and wake detection with natural voice interaction.
Machine learning Β· Computer vision
A custom vision model for friendly-face classification, evaluated against ResNet and CLIP baselines across accuracy, robustness, latency, and real-world image quality.
Product Β· ML and backend systems
A marketplace I co-founded, where I led ML and backend work across trust, commerce, analytics, APIs, and recommendation-oriented product signals.
Rust Β· Developer infrastructure
A Rust and macOS synchronization daemon for AI coding workspaces, built around deterministic templates, atomic writes, and drift detection.
Experience behind the work
Jan 2026 β Apr 2026
Ontario Public Service
Built production RAG and LLM workflows for security and log intelligence using AWS Bedrock, Python, evaluation pipelines, and operational dashboards.
Jan 2022 β Jan 2025
Built high-throughput Go and Python services, REST APIs, and event-driven systems for real-time logistics workflows.
Aug 2021 β May 2024
Led product discovery, ML and backend architecture for a commerce marketplace with trust, fulfilment, analytics, and personalization workflows.
Collaboration
Iβm especially interested in computer vision, applied ML, and systems that have to behave reliably in the real world.