Advisory
Selective AI advisory
Focused guidance for teams making consequential decisions about AI research, evaluation, architecture, and multilingual systems.
What I advise on
Model evaluation
Evaluation design, failure analysis, and evidence-based assessments of whether a model is fit for its intended use.
Research & architecture reviews
Focused review of an AI system, research plan, or technical direction before a team commits significant time and resources.
Multilingual & African-language AI
Data strategy, evaluation, model choices, and community-aware approaches for multilingual products and African contexts.
Technical workshops
Tailored sessions that help research and engineering teams understand a method, assess trade-offs, or align on an AI approach.
How we can work
One-off consultation
A focused conversation to clarify a question, pressure-test a decision, or identify next steps.
Written review
A bounded assessment of a system, dataset, architecture, evaluation plan, or research direction.
Workshop or advisory sprint
A time-boxed engagement with a defined question, audience, deliverable, and end date.
Clear boundaries
Engagements use public or client-owned information and exclude confidential, proprietary, or employer-related work. Every enquiry is subject to conflict review and any required approvals.
Request advisory support
Share your organization or industry, the question, desired output, timing, and any potential conflicts. I accept only a limited number of clearly bounded engagements.