Credentials

A spatial AR interface for building data pipelines through gesture interaction. Users physically snap blocks together to construct pipelines, with haptic feedback confirming each connection as the structure takes shape. The underlying intelligence uses a Support Vector Machine to classify pipeline configurations against preset training data in real time — telling users immediately whether a configuration will succeed or fail, and why. SVM was chosen deliberately over generative AI and reinforcement learning: both alternatives would have introduced complexity and data requirements that worked against the goal of fast, accurate feedback in a structured training environment. The design keeps the AI layer functional and legible rather than generative, which is itself a responsible AI position.

A strategic plan examining how organizations can extend their collective intelligence by integrating AI into core business functions — not replacing human judgment, but augmenting how groups sense, decide, and act together. The work applied superminds theory to identify where AI interventions create compounding value: not at the individual level, but at the organizational level, where coordinated intelligence becomes a structural competitive advantage. The plan connected these frameworks directly to differentiation and growth, making the case for AI as an organizational capability rather than a feature set.

Applied generative adversarial networks to a real-world societal problem. GANs combine generative and discriminative modeling — the generator produces synthetic outputs, the discriminator pushes back — and that adversarial tension is what makes them powerful for problems where realistic synthetic data matters. The project required working within a responsible AI framework that addressed what the course called "AI cancers": bias in training data, explainability of outputs, and resilience to adversarial conditions. The work had to demonstrate not just that the approach was technically feasible, but that it was sustainable and accountable — making the case for the solution as much on ethical grounds as on performance grounds.