Kwesi Walker
Building intelligent systems at the intersection of federated learning, NLP, and reinforcement learning. Pursuing a dual degree — MSc AI in France and BSc Computer Engineering in Ghana. Seeking a 12-month alternance starting September 2026 in research labs or industry AI teams.
Projects
Federated Text Classification with Blockchain Logging
Modular federated learning system using DistilBERT across 3 clients with FedAvg aggregation. Integrated blockchain-inspired logging for traceability. Achieved 81.5% accuracy (+1.7% vs centralized baseline) on AG News.
RL Tennis Agent
PPO agent trained to compete in a Unity tennis environment. Designed a 13-feature observation space and 4-action policy with custom reward shaping and TensorBoard monitoring.
Multi-Task NLP Toolkit
Multi-task NLP system covering sentiment analysis, zero-shot classification, semantic similarity, and extractive QA. Built modular pipelines with interactive Gradio interfaces.
Diabetes Prediction Model
Predictive models on the Pima Indians Diabetes dataset. Applied preprocessing, cross-validation, and compared Logistic Regression vs KNN performance.
VR 3D Word Puzzle
Immersive VR word puzzle game with 3D mechanics, object interaction systems, and spatial audio cues — merging gaming with cognitive challenges.
3D Shooter Game
Physics-based 3D shooter demonstrating player control systems, projectile mechanics, and core gameplay loop design.
Research
Boolean Satisfiability Problem (SAT)
A technical paper on the Boolean Satisfiability (SAT) problem — a foundational NP-complete problem in computational complexity. Analyzed CNF transformations, brute-force vs. optimized solving approaches, and the role of SAT solvers in AI planning, constraint solving, and automated reasoning.
Education
Machine Learning · Deep Learning (CNN, RNN, Transformers) · NLP · Reinforcement Learning · TensorFlow
Linear Algebra · Probability & Statistics · Calculus · Signals & Systems · Numerical Analysis · Data Structures
Pursuing both degrees simultaneously — combining engineering foundations with applied AI research.
Experience
Skills & Stack
Let's build something intelligent.
Seeking a 12-month alternance starting September 2026. Available for research labs, AI startups, and enterprise ML teams — with a focus on scalable ML systems, federated learning, and applied NLP.