Distyly is a crowdsourced biohacking platform: over 240 health topics and 1,200 entries covering dosages, side effects and brands, rated for effectiveness by the people who have actually tried them.

Distyly began with its founder trying to recover from chronic illness and finding that the useful information was scattered. Functional medicine and biohacking knowledge, what people take, at what dose, and what it did to them, moves peer to peer through forums, threads and word of mouth, and formal research on much of it is thin. The brief was to collect that experience in one place, curate it, and signpost the strategies that work, across nutrition, supplements, repurposed medications and lifestyle changes such as exercise.

It had to be somewhere people contribute to, not a directory they read once. Members suggest topics, rate interventions for effectiveness, and flag root causes, investigations and side effects, so the content grows and reorders itself as more people use it. The hard part was keeping it navigable. A library past 248 topics and 1,238 entries only works if a member can find the one thing they came for.

Distyly runs as a web interface on a Firebase backend, with a mobile app in development. Firestore holds the topics, entries, categories and channels and pushes changes straight through to the interface, Firebase Auth covers email and Google sign-in, and Cloud Functions in London do the work sitting behind the content. Algolia powers search across entries, topics, brands, products and the taxonomy underneath them. Each member is issued their own scoped key rather than the whole site sharing one. The design system it is all built on was a collaboration with Base Media.

Most of the care went into the rating system. A topic takes a one-to-ten effectiveness slider, while the five-star control sits on the relations between an entry and a topic, and entries are never rated on their own. Members also build combinations, stacking the entries they take together and rating those as a unit on the same five-star control, with a hash over each combination's documents used to spot stacks that already exist. Scores are held as a Bayesian average, recalculated by Cloud Functions on every interaction and re-aggregated every twelve hours. Behind all of it sits an admin panel where topics, entries, dosages, side effects and resources are managed, and where pasting a product URL is enough to pull the product in, tidy it up and attach it to the right brand.