Assembling @droid, the frontier software development agent. Available to anyone, with any model, in any IDE/terminal.

curl -L app.factory.ai/cli |sh
Factory retweeted
As a data person, I've been waiting for this one!
Announcing our new and improved Analytics, giving engineering leaders transparency into consumption, model efficiency, and adoption, broken down by model and by user across every session in the organization.
2
25
1,267
Factory retweeted
People underestimate not all model-independent providers are the same. If you train a model, you are incentivized to give that model to people. This impacts how transparent your analytics are (so you can prioritize your model). We will never obfuscate this information!
Announcing our new and improved Analytics, giving engineering leaders transparency into consumption, model efficiency, and adoption, broken down by model and by user across every session in the organization.
2
4
46
2,982
Announcing our new and improved Analytics, giving engineering leaders transparency into consumption, model efficiency, and adoption, broken down by model and by user across every session in the organization.
19
11
116
26,409
Agentic software engineering is a growing line item, and platforms earn continued trust by showing exactly where that consumption goes, without placing the burden on engineers.
1
17
1,696
Instead of a single abstracted compute number, Factory reports consumption at the level where decisions get made: which model ran and who ran it. Analytics puts that visibility in one place. Full post: factory.com/news/factory-ana…
1
13
1,001
Factory retweeted
We wanted one place that tells our whole EPD team what's actually true right now - a "Product Operating System". Automations is what powers this. Our ProductOS aggregates key metrics, product taxonomy, customer feedback, internal org-chart, ownership areas, product roadmap, and more - all in one place. Every morning at 8 a droid automation spins up on a cloud computer and reads one skill. that skill kicks off a subagent per playbook, 10 of them. one hits the warehouse, one pulls okrs out of notion, one diffs linear, a few read slack (launch channels, pod channels, support), one counts releases off github. calendar waits for roadmap since every release has to link to a roadmap item. the digest goes last and writes up what changed. Each subagent has to check its own work before writing anything - is the doc count in the range it expects, is the data actually from today. if something's off it doesn't load, logs why, and the site just shows yesterday with a stale banner. One of my favorite parts about automations is how they learn. The first line of the skill is basically "this file is everything you know, assume nothing else exists." so when a run screws up we don't patch code, we write the lesson into the playbook and tomorrow's run picks it up. Our head of product @TaylorTheSavage owns this now and playbook changes go through PRs like everything else. next up is letting it enforce stuff instead of just showing it: no kickoff, no status update in a week, no launch channel, it flags it or just does it.
Custom Automations are now GA in Factory. Describe a recurring workflow, pick a schedule or event trigger, and Droid runs it to the intended result. You choose the model, the machine, and Connectors for each automation.
10
8
84
7,410
factory automations i can't live without as a PM: - daily summary of all new PRs with meaningful user-facing changes - metrics reports that auto-evolve as new features roll out - classify new linear tickets by "will this make me a first-tier competitor", auto-close if no
Custom Automations are now GA in Factory. Describe a recurring workflow, pick a schedule or event trigger, and Droid runs it to the intended result. You choose the model, the machine, and Connectors for each automation.
5
2
58
6,322
Factory retweeted
Problem: my team couldn't keep up with everything that was happening on Slack everyday. Solution: custom droid automation sends slack highlights straight to email and DM. Perfect read over a coffee every morning. Great for engineering, great for knowledge work
Custom Automations are now GA in Factory. Describe a recurring workflow, pick a schedule or event trigger, and Droid runs it to the intended result. You choose the model, the machine, and Connectors for each automation.
1
15
1,300
RT @oliverzhang29: I used to send my agents the same prompts every morning and read the same dashboards. I was the automation. Not anymor…
1
597
Factory retweeted
Automations more than 2xed my weekly PR output! My favorite automations improve our code quality: - Eliminate code smells and dead code - Improve and fix documentation - Fix flaky tests and optimize slow ones Set these up today and improve your code every night while you sleep
Custom Automations are now GA in Factory. Describe a recurring workflow, pick a schedule or event trigger, and Droid runs it to the intended result. You choose the model, the machine, and Connectors for each automation.
2
1
24
1,218
Custom Automations are now GA in Factory. Describe a recurring workflow, pick a schedule or event trigger, and Droid runs it to the intended result. You choose the model, the machine, and Connectors for each automation.
19
14
191
121,479
Pick the model and reasoning level for each automation. Among current Automations users, 30% of all code review automation runs use an open-weight model.
1
1
14
1,444
Factory is a launch partner for the OpenAI B2B Marketplace. Eligible @OpenAI enterprise customers can now apply part of their existing OpenAI commitment toward Factory, driving engineering efficiency and accelerating software development with autonomous AI agents.
19
8
145
28,187
Factory retweeted
Factory CEO on why AGI is already here @matanSF is the co-founder and CEO of @FactoryAI, a $5 billion AI coding startup building Droid agents to automate the engineering process. He tells me how AI is changing the job of a software engineer and why Factory is betting on a future with many competing models. We get into how it balances model performance with token costs and why he thinks companies should avoid depending on one AI provider. We also discuss why he thinks AGI is already here, how Factory hires its own engineers, his view of the open-versus-closed AI debate, competition with companies like Cognition, and when companies should outsource to AI rather than build in-house. Timestamps: 00:00 The AI coding shift 06:10 What makes Factory different 11:32 Funding Factory’s growth 16:17 When AI writes the code 20:25 Hiring in the AI era 28:30 Betting on multiple models 32:49 Is AGI already here? 34:30 AI safety and regulation 39:36 Building a trusted AI business 41:32 The risks of model dependence 44:50 Competition and engineering’s future Links: YouTube: piped.video/IMxtXy2SnH0 Spotify: open.spotify.com/show/3p1KZv… Substack: sources.news/p/factory-ceo-a… Thanks to the show’s premiere sponsors: @Atlassian, @meetgranola, and @mercury.
3
9
42
63,301
Sonnet 5.5 is live in Factory. Some initial observations: - High is a strong default - Checks that the real requirement is met, not just the nearest failing test - Questions explanations carried over from earlier work Try it now: factory.com/
12
5
108
11,338