AGENTIC SYSTEMS · MCP
Agentic systems & MCP
Claude and OpenAI agents that plan, call your tools through MCP, and ship with guardrails, audit trails and rollback, so multi-step work runs without you babysitting it.
ClaudeOpenAIMCPGuardrails
SENIOR AI & DATA SYSTEMS CONSULTANT
Production AI systems. Built to ship.
I help organizations move from AI ambition to governed systems in production, spanning agentic workflows, knowledge infrastructure, and the data platforms beneath them.
10+ years across the Government of Canada, the United Nations, Lyft, Oliver Wyman and major enterprises in North America and the Middle East. Senior enough to define the architecture, hands-on enough to build it.
Ottawa-based. Working with organizations across Canada, the United States and the Middle East.
Experience across
Global technology companies, government, international organizations, financial institutions and major Middle Eastern enterprises. Production delivery inside high-trust, high-scale environments.
Selected engagements. Full write-ups on Selected work.
Capabilities · production, not demos
Three ways to ship.
The systems I design and personally build. Each module includes a working diagram, scope and a path to production.
KNOWLEDGE INFRASTRUCTURE
RAG & retrieval pipelines
Grounded retrieval over organizational knowledge: chunk, embed, hybrid search, rerank and cited answers with provenance, eval and access controls.
Hybrid retrievalCited answers
DATA & AI FOUNDATIONS
Data platforms & cloud foundations
Ingest to governed products: batch and stream pipelines, lakehouse bronze→gold, model rollout, lineage and access, so AI has data it can trust.
AWS · Azure · GCPDatabricksLineage & access
Proof · high-signal study
Featured work
One study sampled from the Work index, with real stack and outcome, context, constraint, and system.
United NationsInternational OrganizationGlobal
United Nations: Annual Report Automation & Knowledge Platform
Turn fragmented institutional knowledge into governed, reviewable reports and reusable AI access.
Architecture
Engagement model · how organizations work with me
From assessment to embedded leadership.
01
Architecture & production readiness assessment
A defined, low-risk first engagement: stakeholder interviews, current-state architecture review, data readiness, use-case scoring, risk and governance, and a prioritized roadmap.
Explore the assessment →02
Architecture & implementation
I lead and build the systems the assessment recommends: production RAG, agentic workflows, data platforms, evaluation frameworks and the cloud architecture beneath them.
See capabilities →03
Fractional technical leadership
Principal-level AI and data leadership without a full-time hire: architecture review, technical direction across executives and engineering teams, and hands-on implementation where it matters.
Explore an engagement →How we work
From assessment to operation.
A pragmatic sequence that keeps governance, eval and handover in scope from day one.
- 01
Assess
Map the problem, data and constraints with stakeholders. Define success, readiness and the shortest path to a production outcome.
- 02
Prototype
Build a working slice over real data: retrieval, tools and prompts, with eval and governance baked in from day one.
- 03
Productionize
Harden for reliability: access controls, observability, latency budgets and fallback paths before autonomous execution.
- 04
Operate
Ship, monitor and iterate with the team. Handover with runbooks and lineage, not black boxes.
Book a working session
Direct booking, no form required
Discuss your AI initiative
30 minutes to talk through your initiative, the data behind it, and the shortest path to a production outcome.
30 minutes to talk through your AI initiative, the data behind it and the shortest path to a production outcome. Direct Calendly link. Email fallback available.