Komlan Thibaut BASSA

Komlan Thibaut BASSA

I like to learn new things.

01

Education

M.S. in AI & Big Data

2022 — 2024

University of Lomé/UTBM (France)

B.S. in Mathematics

2015 — 2018

University of Lomé

auto_stories "Study hard what interests you the most in the most undisciplined, irreverent and original manner possible." — Richard P. Feynman
02

Career Chronicle

Data Scientist

March 2025 — Present

Mitsio Motu

Performed spatial analysis on Geoportal data to provide actionable insights for several ministries. Developed automated data pipelines ensuring continuous database updates, and conducted literature review on estimating deforestation via satellite imagery.

Part-time Lecturer

Feb 2024 — March 2025

University of Lomé

Instructed a course on Data Mining with a focus on information extraction from textual data. Designed curriculum bridging theoretical NLP heuristics with practical tokenization architectures.

Research Internship

Feb 2024 — Nov 2024

Colibri

Conducted a comparative study solving differential equations. Developed Physics-Informed Neural Networks (PINNs) and benchmarked them against the proprietary Hybrid Differential Equation Solvers (HDES) method developed by Colibri researchers.

03

Technical Initiatives

PINNs vs HDES for Differential Equation Solving

Master thesis exploring the comparative effectiveness of Physics-Informed Neural Networks (PINNs) against Hybrid Differential Equation Solvers (HDES) for solving the Flame differential equation. Synthesizes concepts of deep learning approximation with traditional numerical stability constraints.

Research PINNs PyTorch

Geoportal Spatial Analysis & Deforestation Modeling

Built end-to-end data pipelines for the Togolese Geoportal database. Researched and structured experimental frameworks for predicting intensive agriculture deforestation using satellite imagery and convolutional neural architectures.

Spatial Analytics GIS Satellite Imagery
04

Articles

Inquiries & Collaboration