What is Ballet?
Ballet is a workflow automation platform from the team behind Brainfish. Describe a process in plain English, and its coding agent builds the steps, writes the integrations, and runs the work across your CRM, billing, helpdesk, and internal tools. Every workflow becomes reviewable, version-controlled code, so results stay consistent on every run.
Top Features:
- Chat builder: describe the outcome and get a working playbook back in minutes.
- Any API: writes integrations for any service, even undocumented internal systems your team owns.
- Deterministic runs: playbooks compile to code, so identical inputs give identical outputs.
Use Cases:
- Lead routing: match event attendees to the CRM, enrich them, and assign owners.
- Support triage: read tickets, check the order, and draft a reply for review.
- CRM hygiene: find duplicate accounts nightly, merge safe ones, and flag the rest.
Who Can Use Ballet?
- RevOps teams: automate routing, enrichment, and cleanup without waiting for engineering time.
- Support leaders: turn common tickets and deal handovers into drafts people only review.
- Engineering leaders: approve readable code, set roles, and replay any past run.
Pricing
- Workflow review (free): a 30-minute session that scopes one backlog workflow with you.
- Early access (contact sales): design partner onboarding for RevOps, support ops, and growth teams.
- Enterprise (contact sales): roles, approval gates, Git sync, your own models, and replays.
Pros and Cons
Pros:
- Predictable cost: compiled playbooks skip repeat model calls, keeping runs fast and cheap.
- Flexible access: run playbooks from Slack, a Chrome side panel, or Claude.
- Governance: version history, approval gates, and SOC 2 Type II coverage via Brainfish.
Cons:
- No public pricing: you must book a demo before learning what it costs.
- Early stage: still onboarding design partners, so features and terms may change quickly.
- Revenue focus: examples center on sales, support, and finance rather than general work.
FAQs:
1) How is it different from Zapier?
It writes integrations for any API and runs workflows as code, not visual graphs.
2) Do engineers need to build workflows?
No, anyone describes the outcome in chat, and engineers just approve the code.
3) Are results consistent?
Yes, playbooks compile to code, with AI judgment only on steps that need it.
4) Can I use my own model accounts?
Yes, the gateway can point at your existing OpenAI or Anthropic account.
5) How does Brainfish fit in?
Brainfish handles knowledge, this tool handles execution, and it also runs standalone.