An operator-first surface
Every control is reachable without docs. No webhook URLs, no JSON paths — just record what you already do.
The AI screen recorder for repetitive work.
Backed by engineers from
Record,replayandautomateany task you do on repeat

Every flow streams live — per-step logs, inputs, outputs, elapsed time. Retry any failure in one click and audit the diff.
Every control is reachable without docs. No webhook URLs, no JSON paths — just record what you already do.
Flows stay current on every machine the moment you save.
Dropplit keeps the full recording context so flows run correctly even when later steps depend on earlier decisions.
Queue runs offline and let them execute on reconnect — nothing lost when the network drops.
Teammates watch runs live, leave step-level notes, and retry a failure together in one shared thread.
Recordings stay encrypted end-to-end. Share runs with named teammates — never a public link by default.
Mark fields and frames private. The model never sees credentials, DMs, or anything you don't share.
Every save is a checkpoint. Roll back a flow in one click and diff runs.
The Dropplit recorder watches you triage, reply, file, and update — no webhooks, no JSON paths. One keystroke captures a minute of work and compiles the intent underneath.

Transcript, typed text, and clicks become a 14-step workflow with conditions, retries, and branches. Audit it, edit it, roll it back. No black boxes.

OAuth once, then Dropplit runs on our cloud runtime. Native connectors for Gmail, Slack, Sheets, Notion, Calendar, Drive — bring-your-own-tokens for everything else.

Live run feed, per-step logs, one-click retries, signed audit trail. The kind of observability your ops team actually needs on the Monday after a launch.
