Warden

Problem Finders starts one step earlier than most platforms: before solutions, it finds the problems that are actually worth solving. Today, the same problems get solved again and again by different people who never meet, and good questions get lost in feeds and short videos before anyone acts on them. A person with a real problem often has no clear place to post it, and no small team to work on it with. Problem Finders gives that question a place to live, a number attached to it, and a team around it. Anyone can submit a problem with who it affects, what has already been tried and why it fell short, the skills needed, and a measurable first win. Others can support it, save it, discuss it, and join a small team of 3 to 5 people to work on it. Points and a leaderboard reward supporting, joining, submitting, and commenting. The board starts with 12 concrete problems across 9 categories — for example, flood warnings arriving too late for river villages in Bangladesh, ICU bed hunting in Dhaka, vaccine cold-chain failures, undetectable city water leaks, students locked to 2G, and deaf transit riders — each labelled by whether it can be solved by programming, AI, engineering, or biology. It is a working demo in a single file with no build step: browsing, search, filters, sorts, supporting, saving, commenting, team seats, submitting, points, and the leaderboard all work in the browser. Accounts, a server, and saved data across reloads are not built yet — that is the next step, followed by AI help on a pitch, including duplicate detection.