Author: Prasham Shah
May 5, 2026
When I look across the projects I am most drawn to, the common thread is not one stack. It is systems that turn messy information into better decisions.
In AI products, that means retrieval, ranking, summarization, recommendations, and workflow design. A system is only useful if it helps someone do something they already care about. That is why I like building products around research, shopping, and personalized information flows. The output has to reduce friction.
In finance, the same principle shows up as interpretability. Markets are noisy. A model can produce a signal, but the better question is why the signal exists and when it stops being valid. Explainability matters because decisions need trust.
In systems work, I keep coming back to constraints: latency, compute, memory, scaling, and reliability. The best product idea still has to survive the engineering details. That is why coursework around parallel computing, operating systems, and machine learning systems feels connected to everything else I build.
Web apps bring the loop together. The frontend is where a user’s mental model meets the system. If the interface is confusing, the intelligence underneath does not matter. Good products make complex systems feel understandable.
So my technical focus is broad, but not random: AI, finance, GPU and systems thinking, and web software all point toward the same goal. Build systems that understand context, explain themselves, and make real decisions easier.