One chat in. A queryable life out.
There are exactly four steps. You only ever do the first one.
You talk like you would to a friend
No fields, no taxonomies, no "select category". You type what happened. Fyg doesn't care if it's a meal, a workout, a task, a mood, a meeting, or a dollar amount. Whatever you mention, it picks up.
It parses the message into structured records
Behind the scenes, Fyg identifies what kind of thing you mentioned — meal, task, observation — and fills in the relevant fields. It applies units, timestamps, categories, and consistency rules so your data stays clean over months and years.
See it land. Edit if it's wrong. Move on.
Fyg confirms what it captured in plain English. If something looks off, fix it in a follow-up sentence ("actually it was 30 grams, not 13") or click into the entry directly. No "log entry approval" workflow.
Then ask the same chat anything about your data
The chat that captures is the chat that queries. "What did I eat Tuesday?" "Show my mood over the last month." "Did my caffeine intake change after I cut afternoon coffee?" Fyg answers with text, charts, or the underlying entries.
Built on real infrastructure, not a chat-screen demo
Fyg runs on a real PostgreSQL database. Every entry is a row. Every query you ask the chat is backed by SQL — not by an LLM pretending to remember.
The chat layer uses modern frontier models for the natural-language understanding, connected to Fyg’s data tools via the Model Context Protocol. You can read the same data via export or API — your data isn’t stuck behind a chat window.
Limited slots open for the next cohort
Fyg is in closed Beta and we onboard testers in small cohorts to keep feedback tight. Drop your email — we'll reach out when a slot opens.
- Hands-on access to a working product, not a waitlist screenshot
- A direct line to the maker — your feedback shapes the roadmap
- Free during Beta · grandfathered rates if you stay on