Tom Atkinson

Changing the world, one Coworker at a time

[Coworker.ai @coworkerapp]

More agents and models make it harder to see where AI spend goes. Our rebuilt Analytics dashboard shows how Coworker credits are used: agent runs, top agents, and credits by model.

Set credit caps per model, control model access per person, and export to CSV 👇

[Tom Atkinson @T__W__A__]

@coworkerapp let them cook!

[Coworker.ai @coworkerapp]

Code is in beta 🚀

Every model your engineers need, from near-free open models to frontier, with your company’s context already loaded and usage limits you set by model and by person.

It runs right in your terminal, installed with one command 👇

[Michael Saeta @SaetaMicha38987]

We will run that same benchmark for you. Prove it in the data! Cheaper, better, faster!

[Alex Calder @ATCalder]

9x cheaper, 64% faster, preferred on quality 84.5% of the time. No change to your existing workflows.

[Peter Zhou @xtzhou]

Don’t use AI that forgets after every session. Bring your context to any model. Love what @ATCalder and @HeresTheChurch have built!

[Jeff Huber 🇺🇸 @jhuber]

Congrats to @coworkerapp on OM2! The next step for enterprise AI that actually works isn't a bigger model, it's memory.

[Tom Atkinson @T__W__A__]

OM2 by @coworkerapp just dropped

[Tom Atkinson @T__W__A__]

OM2 by @coworkerapp just dropped

[Tom Atkinson @T__W__A__]

@coworkerapp Remember having no org memory and having to remember? Seems like a distant memory

[Tom Atkinson @T__W__A__]

@coworkerapp SO good

[Coworker.ai @coworkerapp]

Today we're launching OM2.

Your AI re-reads your entire company from scratch every time you ask it something.

It’s why more than 50% of your token bill isn't in the answer, but in the search for it.

OM2 gives your AI a permanent memory of your company. 🧠

[Alex Wurts @AlexWurts]

If anyone is curious what a true software factory looks like. Check out what happened to our Principal engineer's velocity after he launched our first factory.

His project was building a pipeline for printing data connectors, not just building one connector at a time.

[Tom Atkinson @T__W__A__]

@coworkerapp

[Tom Atkinson @T__W__A__]

If you're doing any manual step on the left, you're doing it wrong

[Rohan Paul @rohanpaul_ai]

The real value is using AI to reduce work. AI that edits, builds, refreshes, and ships work.

@coworkerapp just launched Open Artifacts, turns AI chat into shareable work products grounded in company data.

It completes the job as a shareable doc, deck, spreadsheet, dashboard, app, or agent.

Coworker is solving the problem where nontechnical people want AI to create serious business artifacts without burning huge token budgets.

It also turns those artifacts into shared, editable work instead of isolated personal files.

[Michael Saeta @SaetaMicha38987]

The ability to create a beautiful dashboard that lists out all of my POCs, cross references our highest Jira tickets, assigns a value to each of those and then takes call sentiment to help us understand priority, is now easy.

But more importantly, it doesn't die with me... I can now create, share, update, and manage this in a way that doesn't burn tokens and require my team to know how to build it themselves. Create, share, update. @coworkerapp just gets more and more powerful.

[Tom Atkinson @T__W__A__]

@HeresTheChurch You had me at cheaper/better context/collaboration/free-trial!

[Bradford Church @HeresTheChurch]

Today's a giant upgrade for Coworker: Open Artifacts.

Our vision for https://t.co/S5OepDyjK5 is to be the best place for any team to get real work done, natively AI enabled for collaborating with both humans and their agents. Now we're unlocking the next phase of that vision.

[Alex Calder @ATCalder]

We kept hearing the same problem from Codex and Cowork users:

We like building stuff, but all that context blows a hole in token budgets.

Open Artifacts routes the work to the right model and context, for 4-5x less cost.

The outputs are genuinely beautiful.