Everything your company’s context is made of.
Your data, the objects your business runs on, the metrics you define once, access rules, skills and apps, in one cloud instead of nine tools. Any agent reads it through one connector; every app runs on it. Each block can be used as it is, installed from the marketplace, or built by your agent on the same SDK we use.
The cloud underneath. Apps on top.
A composable stack: collection, storage and processing run as one cloud; a semantic layer gives every app and agent the same numbers; apps sit on top, whether ours, from the marketplace, or the ones your agent writes: a CRM, lead scoring, finance reporting, marketing analytics. Every agent reads the same context underneath.
Apps & agents
Objects, metrics & serving
Processing
Collection
Managed and ready in every workspace: auth to your tools, a warehouse, sign-in for apps, the agent gateway.
Add a finished app, connector or event source from the marketplace. It lands in your repo as code you can change.
Ask your agent for what doesn’t exist yet: a connector, a model or an app. It builds on the same SDK we do.
Everything is a plugin
Every source, destination and agent on the platform, including our own, runs on the same public SDK. If we haven’t built your connector yet, your agent can write it.
Every source your company has, with no pipelines to babysit.
Pipelines that test themselves before they ship.
Analytical, transactional and state storage on one platform.
The model of your business: the objects it runs on and the metrics it counts.
One endpoint for every agent, and a way to act on the answers.
Nine data tools, replaced by one context layer
A typical company pays for nine data tools and the engineers who glue them together, and the numbers still drift between them. Once agents are how people ask questions, BI, the warehouse and the pipelines feeding them become one thing: your company’s context, in one place every agent and app reads. The records you keep in rented CRMs and back-office tools become objects you own, in the same place.
Before: the patchwork stack
~$250k/yr · 9 contracts · 2 data engineers · T+ 6 months
Typical mid-market list prices.
After: one platform
Collect
Process
Store
Model
Serve
Usage-based · 1 contract · first version in two weeks
Deploy directly from your GitHub
One repo is your whole workspace: sources, transformations, metrics and apps live in a folder structure and deploy on push. Every pull request gets a preview environment with its own branch of the data, then merges to production.
One repo, one workspace
Push to deploy
$ git push origin feature/new-margin-metric
▸ preview environment ready · data branch preview-42
▸ checks passed · metrics diff approved
▸ merged → deployed to production
Preview environments cover code and data alike: a pull request branches your warehouse, not just your repo.
Enterprise, with our engineers
Our engineers bring your data, objects, metrics and access rules into your own workspace and build what your teams need, or migrate what you run today, as code you own. It starts with a fixed-scope pilot.
Self-serve, after launch
Self-serve opens to the waitlist first. You’ll create a workspace, connect your own agent and pay only for what runs.
Your company’s context, ready for every agent.
$ segmentstream workspace create acme --region eu
▸ workspace ready · data, objects, metrics and access rules in one place
▸ connect Claude, ChatGPT or Cursor with one connector
▸ tell your agent: “add our billing API as a source”
Our engineers build the first version on your data in two weeks, at a fixed price.
Or join the waitlist for self-serve:
Pilots are open now. Self-serve opens to the waitlist this fall.