Jev AI Experiment
A 25-profile experiment testing whether Jev can reason over structured symptom data without a local matching engine, local scoring, or a fallback layer.
A growing archive of products and experiments. Each case study covers the problem, product reasoning, scope and trade-offs, engineering decisions, and how I would validate or ship the next version.
A 25-profile experiment testing whether Jev can reason over structured symptom data without a local matching engine, local scoring, or a fallback layer.
A Turkmenistan-focused legal platform that turns plain-language legal problems into grounded research and formal documents, with public law browsing and document verification.
A product experiment for structured debate around education, designed to bring students, educators, parents, researchers, and other stakeholders into the same conversation.
For source code and smaller experiments that do not need a full project page yet, my GitHub is the complete archive.