Policy, regulation and risk, translated into product. 

The platform connected African talent to employers hiring across jurisdictions, where employment, tax and right to work rules differ from country to country.

I researched employment and hiring requirements across markets and translated them into how roles were vetted and labelled: which opportunities were genuinely open to international applicants, which required work authorisation in a specific country, and how that was communicated on each listing.

Members could trust that a role marked remote friendly actually was, which reduced wasted applications and protected the platform's credibility with both talent and employers.

The product uses AI to score French writing and speaking against the TCF exam rubric, and a wrong signal could affect a real immigration or citizenship application.

I designed the scoring experience so AI output is presented as practice guidance rather than official results, with clear disclosure that it is automated, rubric explanations shown alongside every score and human review recommended before high stakes decisions.

Learners get fast, useful feedback while the product stays honest about its limits, which is both the ethical choice and the one that builds durable trust.

An agentic assistant that drafts content in a person's voice creates real risk if it publishes something inaccurate, off brand or non compliant with platform rules.

I prioritised voice matching and a mandatory human approval step over automation features like scheduling, so nothing is published without the author reviewing and approving it first.

The product accelerates drafting without ever removing the person from accountability for what goes out under their name.

Growing with analytics usually means collecting more personal data than a product genuinely needs, which creates regulatory exposure and erodes user trust.

I kept analytics focused on behaviour rather than identity: acquisition sources, drop off points and feature adoption, collected with consent and the minimum data needed to make the decision at hand.

The product improved through data while staying aligned with data protection principles, which matters for a platform whose entire value is trust.

Workforce and transformation projects for large institutions carry confidentiality, regulatory and reputational obligations that shape what can be built and how.

Working inside KPMG Advisory and Deloitte's Talent and Transformation teams, I learned to treat regulation, confidentiality and stakeholder risk as design inputs from day one rather than review gates at the end.

That discipline is now how I build: risk, regulation, trust and user protection are considered at the specification stage, where they are cheapest to get right.