I move between boardrooms, models, and manuscripts, same instinct every time: look closely before you speak.

At Qucoon, a software engineering and consulting firm headquartered in Lagos with global presence in the UK, Dubai, and Kenya, and a multicloud partner across AWS, Microsoft, and Huawei.

I work day to day with AWS Account Managers and Solutions Architects on the startup vertical, including Visa's Cohort 3 & 4 startup accelerators. I sell cloud migration, cost optimization, AI predictive systems, AI agents, AI chatbots, AI evals, and AI-powered credit analytics, including tailored offerings for PAYG hardware financiers and issuer-processor/BaaS platforms, to African startups and fintechs, qualifying and structuring every deal with the SPICED framework.

Recognized in NASA's GenHAX SPOKE Challenge, and part of Cohere Labs' Afri Aya Expedition validating Igbo-language datasets. AWS and Anthropic AI Fluency certified.

The Afri-Aya Project, Cohere Labs: 13+ African languages, 2,400+ culturally-rich images.

My evaluation work spans chatbot quality audits, multi-model comparisons, and structured annotation projects, using an approach shaped by Hamel Husain and Shreya Shankar's work on LLM evals.

The clearest example: a full evaluation of a fintech chatbot, covering a 100-case golden dataset, failures mapped across a 24-category taxonomy, and outputs compared across four LLMs.

100-case golden dataset · 24-category failure taxonomy · four-model comparison · Cohen's Kappa validation · multi-turn testing

Not posted here. Please request it directly from me by email.

Philosophical essays, written in the spirit of Baldwin. More at the Substack.