Martin Uwakwe — applied AI, built & governed
I'm an IAPP-certified AI Governance Professional (AIGP) who ships real consumer AI products people use — and runs the evaluation, fairness, and risk work that keeps high-stakes machine learning trustworthy. Most people do one of those. I do both.
aigp certified · nist ai rmf · shipping consumer ai · auditable by design
build // govern
Featured — where building meets governance
ekiti.mundx.com
An AI reads every result form the electoral commission uploads, extracts the figures, and runs validation checks — counting only what passes, with every number traceable back to its source document. Non-partisan by design. It's the clearest proof of how I work: a system whose entire purpose is to be auditable, shipped into a high-stakes civic moment.
method: extract → validate → count only what passes · property: traceable, verifiable, non-partisan
visit ekiti.mundx.com →Consumer AI and data products taken end to end — problem, pipeline, frontend, launch, and real users on the other side.
An LLM pipeline — embeddings, retrieval, structured output — that reads thousands of immigration appeal decisions and flags the patterns likely to trigger a Request for Evidence before a petition is filed.
status: shipped · users: 1,200+ · cases: 3,400+
casereviewer.ai →A full data product — scraping to Postgres to React — turning millions of Department of Labor disclosure records into a fast, searchable salary and sponsorship tool. Grown with zero marketing spend.
status: shipped · monthly users: ~18k · searches: 60k+
h1bdata.us →AI-powered exam prep — adaptive practice and instant, explained feedback that helps students learn from every question instead of just scoring it.
status: live · adaptive practice & instant feedback
examchef.ng →The assurance side of AI — evaluation, fairness, and risk for machine learning that real decisions depend on.
I own the full AI/ML lifecycle for clinical and operational models in a regulated healthcare environment: running model evaluation and testing protocols, conducting algorithmic bias and fairness assessments, and mapping model risk and internal controls to the NIST AI Risk Management Framework alongside compliance and legal teams. I've built 15+ production dashboards supporting planning across 10,000+ locations — with the data-quality and validation controls that high-stakes decisions demand.
scope: aigp · nist ai rmf · bias & fairness assessment · model validation & testing
Designing and managing AI solutions with safety and ethics built in from the start — across finance, healthcare, retail, and supply chain.
focus: responsible deployment · risk & controls
Published work on carbon-credit monitoring and prediction in a smart factory using explainable AI and data analytics — transparency as a first-class property of a model, not an afterthought.
theme: explainability · auditable ML
Judged campus rounds of the world's largest student social-entrepreneurship competition ($1M seed prize).
Organised two editions — content curation, community building, and cross-functional event production.
Years spent making complex careers legible to the most exacting evaluators there are — the deep domain expertise that makes the AI products defensible, kept here as context rather than headline.
The intersection most people can't write from: someone who both builds AI and governs it, working in talent and mobility. Most of my writing is published on Medium.
Read on Medium →What the ekiti tracker taught me about designing systems to be verifiable from day one.
Retrieval, structured output, and the guardrails that matter when the stakes are someone's case.
Mapping model risk to NIST AI RMF in a regulated, high-consequence setting.
Turning millions of government records into tools that change real outcomes.
Get in touch
Open to AI governance roles, advisory work, and collaborations at the intersection of building and assurance.