Pinned Tweet
Today we're launching Ballet.
The way ops does engineering work without engineers. Describe any workflow in plain language. Ballet writes the integration, runs it, and shows you every step.
here's 6 playbooks you can run in Ballet today.
each one is a job ops teams told us eats their week:
All 6 run on a file or a few URLs. Nothing to connect.
Start free: bit.ly/4jgCr4R
6
Ballet is now open to everyone.
Today, you can explore playbooks for Marketing Operations, RevOps, and CX Ops.
Describe a workflow and let Ballet handle things like:
→ Cleaning, enriching, and scoring event lists
→ Auditing CRM data and prioritizing fixes
→ Identifying accounts at risk of renewal
Completely free to try.
It costs nothing to find out if it works for you.
Go see what you can build with it: hubs.ly/Q04Fk8tY0
Here's what Ops teams want, instead of another random AI tool in the stack:
The 5 jobs they repeat every single week, running on their own, with the same steps every time, a record of what happened, and a way to undo it.
Surprise, we're launching self sign-up for this on Monday.
Try it for free and let me know what you build.
5/ Like AI, with the reliability of code. Early access now available, with select teams already in production.
1
82
6/ Learn more: ballet.dev/
71
Daniel Kimber retweeted
AI support that doesn't know your product doesn't deflect tickets. It creates them.
A bigger model won't fix that. A better knowledge layer will.
This month, that knowledge layer shows up in three more places your team already works.
→ Brainfish MCP for @AnthropicAI's Claude
Stop re-briefing Claude every session. Drafts, answers, and conversation guides grounded in what your company actually knows — instantly. Hours back per week, per person.
→ @MicrosoftTeams integration
Every internal question answered in-thread, with citations. New hires ramp in days. Senior people stop being human help desks. Your most knowledgeable teammate is now everyone.
→ @Zendesk live agent handoff
Customers stop repeating themselves. Agents pick up with the full conversation already in front of them. Faster resolutions, higher CSAT, fewer dead ends.
If your AI support isn't pulling its weight, the knowledge underneath it is usually why.
Watch the video ↓
Daniel Kimber retweeted
Everyone's shipping AI agents.
95% of those agents fail.
That's an MIT stat, not ours.
May 7, @_dankimber talks about what the 5% are doing to succeed.
Daniel Kimber retweeted
hosted a Women in AI event in SLC yesterday.
ended up with room full of educated women passionate about AI
ready to share their ideas, what’s changing, and where it’s going
plus we had fish bowl punch (important)
seattle, you’re next.



