Sumon works at the intersection of Software Engineering and AI, focusing on design and verification of fairness and safety of ML-enabled systems. He worked on fairness verification of neural networks and modular reasoning of societal bias of predictive decision-making. Currently, he is investigating long-term fairness issues that are exhibited through the interaction between the system and environment. He is also working on the DARPA project – Verified Security & Performance Enhancement of Large Legacy Software (V-SPELLS), to model and verify safety of Unmanned Aerial Vehicles (UAV).