We use essential cookies to run the site and optional analytics to improve it. You can accept all or keep only what's essential.
AI safety researcher — alignment & evaluation
London, GB · speaks EN, FR
AI safety researcher focused on evaluation, red-teaming, and alignment of large models. Frequent conference speaker across Europe and North America.
Evaluating deceptive alignment (NeurIPS 2024)
Publication
Interviewed by the Financial Times on AI risk
Media Mention
Why this expert ranks the way they do — the engine is deterministic and fully explainable.
Verification level 6 of 6.
Accepting requests · not marked available this week · typically replies in ~24h (faster replies earn the remainder)
5.0★ across 10 reviews.
Earns up to 8 pts when a query requires a language you list.
Query-independent trust — used to break ties between equal match scores.
via reviewer
via reviewer
Evidence reviewed
via reviewer
via reviewer
via reviewer
★★★★★
— Jordan Buyer
By semantic similarity (vector search)
Machine learning engineer — recommendation & ranking
Clinical research & oncology trial design
Fintech fraud & payments risk consultant
Venture due diligence & startup growth advisor
Pick a service on this page and request a booking (or send an inquiry first). Once Okafor accepts, you pay into escrow and the session is confirmed.
Prices are listed per service — no hidden fees for buyers. Payment is held in escrow when you book and released to the expert only after the engagement completes.
Escrow protects you: if a booking is cancelled or declined, the payment is refunded. See the Refund Policy for details. Refund Policy
Dr. Amara Okafor typically replies within ~24 hours. You'll be notified as soon as they respond.
Okafor works in: EN, FR.
Every expert carries a reviewer-assigned trust level from L0 (self-declared) to L6 (board certified), based on verified evidence — publications, credentials, affiliations. Levels are never self-assigned, and the platform blocks any grant whose minimum evidence isn't on file.
Trust level 6 of 6.
5.0★ across 10 reviews (full confidence at 5+).
2 verified proof items (5 = full points).
1 completed engagement (5 = full points).
Verification, availability, reviews and evidence (44 pts) are this expert's standing on every query. Topic relevance (36), semantic fit (6) and location (6) are scored per query; language (8) depends on the query's requirement.
Yes — send a free inquiry from this page, or ask the AI Twin (when enabled) which answers only from Okafor's approved material.