Present
- Principal Developer at Monks
Hello! This is Jeff,
a Software and Product Developer building for the AI era.
Focused on solving problems and improving creative workflows, I have developed and deployed many production-scale applications.
Toronto, ON
Project Aria
Shipped global web platform work for Meta's research and developer-facing product surfaces.
MirrorMe3D
Designed a WebGL 3D analysis platform for pre-operative facial scans and clinical workflows.
ARCPX
Modular agentic platform for composing AI workflows with a DAG engine and persistent memory.
Monksflow
Led agent tooling and context-engineering systems to improve observability and creative pipelines.
The agent problem nobody talks about — why serious workflows benefit from decomposing monolithic agents into small, inspectable DAG steps you can version, audit, and re-run independently.
A decoupled typeahead engine built on ternary search trees — O(log n) prefix search without HTTP round-trips, packaged as an npm module developers can run on the front-end or behind a cached service.
ARCPX is a modular agentic platform where users compose agents to fit their workflow — an asynchronous DAG engine, configurable nodes, and a persistent memory layer. As founding engineer, I own the full lifecycle from technical discovery and system design through production deployment on cloud infrastructure.
Built enterprise agentic applications on AWS with a hybrid retrieval stack — vector search paired with a custom knowledge graph over unstructured data. The architecture improved response accuracy by 20% over vector-only baselines and gave the system a real-time, context-aware knowledge layer suitable for production workloads.
Re-engineered Meta's content management infrastructure, spearheading the migration of complex localized datasets from legacy systems. This architectural overhaul established a scalable foundation for global platform development and multi-region deployment across WhatsApp.com and related properties.
Spearheaded technical development for a humanoid robot featured on stage at AWS re:Invent 2023 — integrating real-time perception, orchestration, and on-device inference into a live demo environment built for a global audience.