I architect decision systems, not just products.
Product leader specializing in OBCE methodology: engineering the context environment where optimal outcomes become inevitable. I build the constraints, traces, and verification systems that make complex AI-native products succeed.
The Four Laws I Operate By
These principles govern every system I architect. They're not philosophy — they're operational constraints.
Law of Outcome Primacy
"The Outcome Must Be Defined Before the Context." Every prompt, every system, every trace must advance a defined Manifest. Context is fuel — if you don't know the destination, you can't choose the right fuel.
Law of Context Sanctity
"The Signal-to-Noise Ratio Determines Intelligence." The context window is finite. Irrelevant context isn't just waste — it's noise that degrades reasoning. Every piece of data must fight for its place.
Law of Constraint
"Constraint is Liability Validation." An unconstrained model is a hallucination machine. Performance is maximized by aggressively reducing the search space. Negative constraints often matter more than positive instructions.
Law of Verification
"Trust is not a Metric. Verification Is." If the outcome cannot be verified, the context is invalid. Every system must output in a format that deterministic validation can check.
The Context Graph
OBCE-Centric Capabilities
Context Architecture
Designing the knowledge graphs, constraint systems, and decision traces that make AI systems predictable and auditable.
Outcome Engineering
Defining verifiable outcomes before system design. Building the validation mechanisms that prove success.
Model Orchestration
Routing decisions to the right model based on task complexity, cost constraints, and historical trace data.
Systems Built on OBCE Principles
JobEasy professional knowledge graph
Graph-first domain model with embedded decision traces. Outcome: explainable matching with provenance tracking and temporal context awareness.
Robotics abstraction framework
Hardware-agnostic integration with aggressive negative constraints. Outcome: 70% reduction in platform lock-in across ROS2, MuJoCo, IsaacSim variants.
Prompt-engineered UX generation system
Reusable context templates with verification hooks and lineage tracking. Outcome: consistent quality, reduced overhead, full decision traceability.
AI-generated t-shirt commerce pipeline
Automated creative generation with commercial validation constraints. Outcome: design-to-market workflow with verifiable quality gates.
Camping platform disruption framework
South African market-entry with defined outcome constraints. Outcome: defensible wedges with verifiable go/no-go criteria.
AI coding platform benchmarking system
Platform-thinking analysis with traceable evaluation criteria. Outcome: positioning assessment with auditable scoring methodology.
PAIA compliance architecture
South African regulatory framework with risk mitigation constraints. Outcome: executive-level governance with verifiable compliance trails.
AI model pricing comparative analysis
Unit economics verification with founder-level resource constraints. Outcome: cost-optimized routing with measurable efficiency gains.
How We Can Work Together
OBCE defines three certification levels. I operate as a Level 3 Governor, but engage at the level your organization needs.
Pass audit scoring
Design constraints
Organizational world models
Define the outcome. Engineer the context.
I'm open to engagements where the challenge requires Governor-level thinking: context architecture, decision trace systems, or organizational AI strategy.