Twitxy Saved

Ivan Burazin

Verified account Daytona

@ivanburazin

building @daytonaio
San Francisco daytona.com Joined January 2009
1.6K Following 23K Followers 14K Posts

Latest posts 5

Pinned post

Ivan Burazin Verified account @ivanburazin
We just crossed $100M in run rate.

Four quarters ago, we were at $1M. Then this year alone, we grew 37.5x.

Not long before that, we did a hard pivot- handed our old customers to competitors, rebuilt from scratch, and relaunched with a few months of runway left.

The bet was: models think on GPUs but work happens on CPUs.

So every agent that writes code, opens a browser, or runs an RL job needs a composable computer that's isolated, stateful, and ready in milliseconds.

$1M to $100M in about a year puts us in the same cohort as the likes of Anthropic and Cursor. It’s humbling to share that lane with companies we’ve learned so much from.
0:22
193 replies 82 reposts 1.3K likes 390K views View on X
Ivan Burazin Verified account @ivanburazin
We don’t optimize for compute or API costs when building new features

If it takes off, then you solve for it. It's an engineering problem.

The first and biggest goal is to show customers the magic and create value for them.
4 replies 0 reposts 14 likes 1.6K views View on X
Ivan Burazin Verified account @ivanburazin
You can feel the CPU shortage starting.

You can still spin up a single EC2 instance.

But try ordering 200, 500, or 1,000 servers from smaller providers and they’re suddenly asking for upfront payments.
8 replies 5 reposts 108 likes 9.1K views View on X
Ivan Burazin Verified account @ivanburazin
Documentation is your product surface when you’re API-first.

If the product is an API, what do your customers actually interact with? The docs. Your logo lives in the documentation. That’s where people form impressions of your company.

Three years ago, nobody wrote good docs. Now AI writes decent ones.

The problem largely solved itself. But the underlying point still stands.
7 replies 1 reposts 23 likes 1.7K views View on X
Ivan Burazin Verified account @ivanburazin
We resemble Stripe more than we resemble AWS.

When you use AWS, you’re deploying to it. Dashboards, Terraform scripts, Helm charts, clicking through things, etc. When you use Stripe, you integrate it with your app. It happens when your customer does something.

That’s what @daytonaio looks like.

- Your agent needs a computer
- it calls the API and the sandbox spins up.
- agent does the work and sandbox spins down

You never touched anything.

We see usage go up when customers are doing well. So we’re basically a proxy for how your company is actually growing.