About

Alimoudine Idrissou

AI Researcher in Mathematical Reasoning, LLM Evaluation, and Scientific Benchmarks

Independent Researcher · Dakar, Senegal

Biography

Alimoudine Idrissou is an independent AI researcher and software engineer based in Dakar, Senegal. His work focuses on mathematical reasoning, language-model evaluation, scientific multimodal benchmarks, and reliable evaluation methodology. He has designed and hand-verified original olympiad-level mathematics problems and scientific evaluation tasks for frontier AI systems. His background combines mathematics, computer science, software engineering, and mathematics education. He holds an M.S. in Computer Science from ESMT Dakar and a B.S. in Mathematics and Computer Science from Université d'Abomey-Calavi.

Before turning to evaluation research, he taught secondary mathematics in Benin and contributed to curriculum material, and he has since built production software systems across several organisations. That combination — teaching, mathematics, and engineering — shapes how he approaches evaluation: a task is only useful if someone can check it, and a system is only useful if it runs.

Illustrated portrait of Alimoudine Idrissou

Research interests

Mathematical reasoning and proof auditing
Studying whether reasoning systems can produce, inspect, and repair mathematically valid arguments rather than merely plausible final answers. The interesting cases are not arithmetic slips but arguments that read well and quietly depend on something never established.
Evaluation validity and benchmark design
Designing original, hand-verified tasks and evaluation procedures that expose model limitations while controlling for ambiguity, contamination, and weak answer checking. Difficulty and validity are separate properties, and a hard benchmark can still measure the wrong thing.
Scientific and multimodal reasoning
Evaluating how models integrate textual, visual, and quantitative evidence in scientific problem-solving settings, including tasks constructed so that neither the image nor the text alone is sufficient.
Multilingual and low-resource AI
Building and evaluating systems for African languages and contexts that are often poorly represented in mainstream AI benchmarks, where the gap is both in model quality and in the measurement apparatus itself.

Research statement, current work, and methodology

Experience

Research

2026

Research Consultant — Turing

Contract research work designing original evaluation material for frontier language and multimodal systems, aimed at multi-step reasoning that standard benchmark items do not reliably require.

Wrote and independently hand-verified original olympiad-level mathematics problems spanning geometry, number theory, and combinatorics
Designed original scientific tasks requiring reasoning over visual and textual material together
Delivered each item complete: problem statement typeset in LaTeX, ideal response, and explicit evaluation criteria
Assessed how frontier language and multimodal systems handle difficult multi-step reasoning

This material was produced for Turing and remains theirs. Problem statements, ideal responses, evaluation criteria, and model outputs are confidential and are not published here.

2020

Independent Research — Université d'Abomey-Calavi · Cotonou, Benin

Investigated applications of nth-order probability in normal distribution theory.

Applications of nth-order probability in normal distribution theory

Engineering, teaching, and other

2026

Software Engineer — Turing

Refined generated programming problems to competitive-programming standards and validated the solutions against them.

Refined generated programming problems to precise competitive-programming standards
Built and validated multi-step solution strategies
Compared brute-force and optimal algorithms for correctness and efficiency

2025 – 2026

AI/ML Software Engineer — AJIT

Built and shipped AI and machine-learning components within production software systems.

AI and machine-learning system development
Integration of model-backed features into production services

2024 – 2025

Co-Founder — Bindali

Co-founded Bindali and led its technical direction, including the applied AI product work carried out under it.

Technical direction
Applied AI product development

2022 –

Contributor — RJBD

Contributed to a digital platform connecting the Beninese diaspora, including work on secure data handling, and supported junior developers on the project.

Platform engineering
Mentoring junior developers

2022 – 2025

Full Stack Software Developer — Zone01 Dakar · Dakar, Senegal

Built full-stack systems across a peer-to-peer software engineering programme, spanning backend services, algorithms, and applied AI projects.

Full-stack application development across several languages and stacks
Algorithms and systems work, including graph search and simulation projects
Applied AI and machine-learning project work

2019 – 2021

Mathematics Teacher and Curriculum Contributor — Secondary schools, Benin · Benin

Taught secondary mathematics and contributed to curriculum material.

Secondary mathematics instruction
Curriculum development contribution

Education

2023

M.S. in Computer Science — Digital Transformation

ESMT Dakar · Dakar, Senegal

Studied on a Benin government international scholarship.

2021

B.S. in Mathematics and Computer Science

Université d'Abomey-Calavi · Cotonou, Benin

Faculty of Science and Technology (FAST). Studied concurrently with ENEAM.

2021

B.S. in Planning

ENEAM, Université d'Abomey-Calavi · Cotonou, Benin

Studied concurrently with the mathematics and computer science programme.

Honors and scholarships

2025

First Place — EdTech

African Institute for Mathematical Sciences (AIMS) · Hackathon de l'Innovation Scientifique

Awarded for GandaLab / SymLab.

2025

Entrepreneurship Award

Force-N — UNCHK × Mastercard Foundation

Recognition for CVBABA.

—

National Government International Scholarship

Republic of Benin

Two years of graduate study at ESMT Dakar.

—

National Government Scholarship

Republic of Benin

Three years at the Faculty of Science and Technology, Université d'Abomey-Calavi.

Technical toolkit

Mathematics

Olympiad geometry, number theory, combinatorics
Probability and statistics
Proof construction and verification

Evaluation & NLP

Benchmark and task design
Reference-solution construction
Failure taxonomies and error analysis
Transformers, fine-tuning, named entity recognition

Programming

Python, TypeScript, Go, Rust
PyTorch, Transformers, scikit-learn, spaCy
Next.js, React, Node.js
Docker, CI/CD, cloud deployment

Languages

French: Native / bilingual
Foodo: Native
English: Professional
Fon: Fluent
Dendi: Fluent
Arabic: Elementary

Research correspondence

I read messages about evaluation methodology, mathematical reasoning, proof auditing, benchmark design, and multilingual NLP. Email is the reliable channel.

alimoudineidrissou@gmail.com

View CV