Ibrahim Denis Fofanah
Data Scientist & AI Researcher
Data scientist and AI researcher at Pace University. I coined Artificial Frictional Unemployment, and built the first machine learning model for crop yield prediction in Sierra Leone. Author of Understanding Agentic AI. I write about agentic systems and applied ML — with a bias toward what actually works, and who gets left out when it doesn't.
LangGraph vs AutoGen vs CrewAI: The 2026 Comparison Is Already Out of Date
The popular three-way comparison is stale: AutoGen is now maintenance-only, folded into the Microsoft Agent Framework (1.0, April 2026). The real live choice is LangGraph vs CrewAI, and the honest first question is whether you need a framework at all.
Ibrahim Denis Fofanah·Sep 11, 2026·7 min
Your Pipeline Passed. Your Dashboard Is Still Wrong.
A green pipeline only proves the code ran. This practical guide shows how contracts, freshness checks, semantic tests, lineage, and safe promotion catch schema drift before it corrupts a dashboard.
Ibrahim Denis Fofanah·Sep 11, 2026·7 min
Claude Formalized Fermat’s Last Theorem. It Did Not Discover a New Proof.
Claude generated a 13-million-line Lean formalization of Fermat’s Last Theorem in 11 days. The real advance is not a new proof—it is verification throughput, agent scaffolding, and a public artifact that exposes exactly what the kernel checked.
Ibrahim Denis Fofanah·Sep 8, 2026·9 min
AI in Africa in 2026: The Real State of Play
Africa has 17% of the world's population and 0.6% of its data centre capacity, and even if every announced project is built, that share will not improve. A practitioner's map of the money, the compute problem, and what it means for what you can actually build.
Ibrahim Denis Fofanah·Sep 8, 2026·6 min
A Portfolio Project Without Tests Is Just a Screenshot
A polished dashboard shows what you built. A tested, rerunnable project shows why the result should be trusted. Here is the minimum audit trail every serious data portfolio needs.
Ibrahim Denis Fofanah·Sep 8, 2026·8 min
Your Train-Test Split Is Leaking the Future
A random split can make a forecasting model look excellent by training on records that occur after its test rows. Here is how to build a time-aware evaluation that rehearses production instead of leaking the future.
Ibrahim Denis Fofanah·Sep 7, 2026·10 min
Same Model, Same Benchmark: 54.8% or 99.9% Depending on the Harness
GPT-6 Astra scored 54.82% or 99.95% on ARC-AGI-3 at the same reasoning level. The model did not change; the evaluation harness did. That gap is a lesson for anyone measuring agents.
Ibrahim Denis Fofanah·Sep 6, 2026·9 min
Reasoning Models Explained: What Test-Time Compute Actually Costs You
Letting a model think longer is a real capability (DeepSeek-R1 lifted AIME accuracy from 15.6% to 71.0%) and a real bill you pay in tokens and latency on every query. When it is worth paying, and when it is not.
Ibrahim Denis Fofanah·Aug 11, 2026·7 min
Serving a Model: Endpoint, Batch, or Embedded?
Three ways to serve a model, endpoint, batch, or embedded, and the one question that decides which. Most people default to the always-on endpoint. It is the priciest and most fragile option, and usually the wrong one.
Ibrahim Denis Fofanah·Aug 8, 2026·7 min
A 125M-Parameter Model for 11 Languages Is Not a Small Story
UCT researchers built an open corpus and a 125M-parameter decoder-only model covering all 11 official South African languages, nine of them low-resource. The corpus is the real contribution, and the size tells you why.
Ibrahim Denis Fofanah·Jul 31, 2026·6 min
The Broken Ruler: Why "LLMs Now Translate African Languages" Doesn't Mean What You Think
A new study on Hausa and Fongbe translation found that for Hausa, human evaluators preferred a system that no automatic metric ranked first — and BERTScore collapsed entirely. Before you trust a rising leaderboard score in any African language, check the ruler.
Ibrahim Denis Fofanah·Jul 28, 2026·7 min
Rwanda Is Betting It Can Lead African AI Without Owning the Compute
Rwanda is first in Africa on AI policy, agency and summits, but the GPUs are going to South Africa, Egypt, Kenya, Morocco and Nigeria. Inside a small country's bet on governing AI instead of hosting it.
Ibrahim Denis Fofanah·Jul 28, 2026·7 min