I am a Fractional Head of AI & Systems Architect with 18+ years of delivery leadership across tier-1 institutions (KBC, Belfius, Evonik). I help engineering organizations move AI pilots into disciplined production - cutting runaway model spend by 74% while enforcing deterministic repository governance.
Available for a dedicated primary anchor mandate (3-4 days/week) or selective fractional advisory retainers (1-2 days/week), structured with strict organizational boundaries and dedicated focus.
A pragmatic engineering framework to insulate codebase assets from vendor drift.
I design the structural body that makes the AI brain useful.
Right now, the industry is drowning in "Semantic Debt," the cost of burying critical business logic inside natural language prompts. When model providers update their weights, these prompts break. Your AI strategy drifts.
My methodology focuses on building the AI Skeleton. By separating reasoning (the model) from logic (deterministic schemas and code), I ensure absolute portability. If a superior LLM is released tomorrow, you don't rebuild your system, you simply swap the engine.
As the creator of the Sprawl CLI and co-shaper of the lean vision behind the open-source Atomic Agents framework, I bring pre-built technical playbooks to my clients, establishing repository boundary governance and GitOps rulesets on Day 1.
A history of delivering zero-overhead control systems and developer environments in high-stakes industries.
Commercial Impact: Stabilized core multi-tenant platform architecture, eliminated runaway AI token burn by 74%, and co-delivered developer governance rollouts across multiple client organizations totaling 60+ developers.
custom/) with Node.js AsyncLocalStorage and a 17-suite test harness (211 unit tests running in ~2s).Re-engineered the Digital Production department into high-velocity developer pods, achieving a 35% delivery turnaround improvement across client deliverables. Centralized repository knowledge via local rulesets, syncing workspace parameters globally to ensure deterministic environment parity and local containment safety.
Designed and validated high-impact AI Blueprints for CAPA Investigations and Network Incident Control. Helped shape the vision and lean software principles for the open-source Atomic Agents framework to establish a schema-first methodology for deterministic agentic pipelines.
Directed 13 specialists across 2 departments (engineering and design) in the digital transformation of global predictive analytics. Refactored complex, high-overhead frameworks into lightweight Vanilla Web Component control panels, establishing the PHY-UI system for distributed teams, cutting feature cycles by 25% and maintenance costs by 30%.
Standardized frontend architecture across departments; engineered company-wide UI component libraries using Vanilla Web Components and containerized CI/CD test pipelines complying with strict banking security regulations.
Spearheaded the architecture for KBC Touch, Invest, and Appointments mobile platforms. Served on KBC's Architecture Review Board (ARB) and chaired the Orchid Review Board, governing mission-critical digital banking assets across distributed teams.
Targeted Advisory. Zero Bloat. Strategic mandates for organizations looking to scale AI-native systems securely.
Available for a dedicated primary anchor mandate (3-4 days/week) or selective fractional advisory retainers (1-2 days/week). I do not sell open-ended developer hours - I deliver exact system blueprints, configure governed repository boundaries, upskill core engineering leads, and construct strict, enforceable AI governance policies.
Hands-on engineering methodology and review board architecture for teams adopting AI coding agents. I work directly alongside your engineering leads as a Systems Architect to establish company standards, manage token burn, and maintain velocity without architectural chaos.
Engagement: 10 sprint sessions across 4-5 weeks working on active backlogs.
Dedicated primary anchor mandate or selective advisory retainers for systems architecture, Architecture Review Board (ARB) steering, and zero-drift LLMOps.
Commercial Terms: Selective 1-4 days / week advisory mandates.
The governance tooling and operational delivery model I deploy into production environments.
An Applied R&D Lab engineering zero-dependency GitOps governance and multi-IDE rule synchronization for autonomous coding agents, and authoring the open-source Atomic Agentic Fabric (AAF) specification. Also co-creator of the open-source Atomic Agents framework.
Applied R&D Lab / Agent GovernanceAn end-to-end advisory delivery model combining a structured 5-phase governance protocol, hands-on developer pairing workshops, and bespoke stack modernization. From auditing tool fragmentation and formulating version-controlled Company DNA rulesets to live codebase migrations and production telemetry, I transition engineering fleets from unmonitored experiments into governed, production-grade AI velocity.
Framework / End-to-End DeliveryInsights on architecture, engineering leadership, and building at scale.
A CTO operating playbook for scaling an AI-native engineering organization, detailing autonomous delivery pods, the 72-hour context onboarding rule, and state-driven repository governance.
Read Article →
A real-world advisory case study of an enterprise implementing a 3-tier version-controlled AI rule engine across 200 developers to eliminate prompt drift and secure the workstation boundary.
Read Article →
The origin story of the Atomic Agentic Fabric (AAF) specification, detailing how enterprise shadow AI and crippled in-house LLM wrappers forced the creation of a file-based, role-governed agent OS.
Read Article →Stop running ad-hoc pilot projects. Schedule an executive briefing to audit your agentic roadmap, enforce spend caps, or discuss a fractional leadership mandate.