OKEMA was founded on a simple conviction: enterprises rarely struggle because they lack data. They struggle because they lack systems that consistently recommend the best decision.
Research and delivery, held by two people.
After years building production AI systems at Microsoft and conducting research in causal inference, reinforcement learning, and recommender systems, one pattern kept repeating: organizations had the data, the dashboards, and the models, but the decisions still came down to intuition, manual processes, and static reports.
A churn score doesn't save a customer. A fraud flag doesn't close a case. A forecast doesn't set a price. Value is created at the moment of the decision, and that moment was being left to chance.
"Predictions don't create value. Decisions do. OKEMA exists to build that decision layer."
OKEMA helps insurers, banks, and telecom operators move from data visibility to decision intelligence, turning the systems they already run into ones that recommend, act, and learn. We start with a focused roadmap and grow into the decision layer for the enterprise.
Decision systems fail in the gap between a model that works and a system that runs. OKEMA is deliberately structured so both sides have a named owner.
Founder & CEO · Applied Science
Decides what the system should recommend: causal inference, contextual bandits, treatment-effect modeling, and the applied-science choices that turn enterprise data into decisions operators can act on.
Engineering & Delivery · Lomé, Togo
Builds the systems that carry a recommendation from model output to the screen where someone acts on it: the services, data pipelines, integrations and operator interfaces that make a decision layer something an enterprise can actually run.
8+ years of applied AI and ML across causal inference, reinforcement learning, recommender systems, predictive and prescriptive modeling, and responsible AI, deployed into real products at enterprise scale.
Built and deployed ML systems into products; LTV impact via causal and language models.
Recommender systems and experimentation at consumer scale.
Production ML pipelines for marketplace decisions.
Causal inference, contextual bandits, and health-intervention systems.
Reflects the founder's prior employment and research, not OKEMA customer relationships.
Whether you're exploring AI or ready to act, the first step is a focused conversation about your highest-value decisions.
Book a call →