recursiv
02 · Build

Agents doing real work, on your real data.

Agents that read your systems, do the work, check it and hand off to a person when your rules say so. On any model, with governance built in.

Work queue · todaySAMPLE
  • AReconcile 214 card transactionsBank feed · ledgerDone · checked
  • AReply to a customer about a late orderSupport inbox · ordersDone · checked
  • AUpdate 38 CRM records from this week’s emailEmail · CRMRunning
  • AInvoice over the approval limitERP · policySent to a person
  • PApproved by the finance leadHand-offApproved
  • ADraft three supplier renewalsContracts · emailQueued
1,284 tasks this week97.8% right6 sent to a person
Anatomy of an agent

Everything an agent needs. One platform.

What triggers it, what it knows, what it can touch, what it does, and the runtime and rules underneath all of it.

TriggersSCHEDULES · EVENTSYour dataCONTEXT · FILESToolsCONNECTORS · APIsActionsIN YOUR SYSTEMSHand-offTO A PERSONRecordEVERY STEPAny modelSWITCH PER TASKInstructionsHOW YOU WORKAgentMEMORY · PLANNING · TOOLSRuntimeSANDBOXES · MEMORY · SCHEDULING · HOSTING · SCALINGGovernancePOLICY ENGINE · PERMISSIONS · APPROVALS · CHECKS ON EVERY RESULT · AUDIT LOG
The parts

What Build gives you.

You bring the work and the rules. Build runs the rest.

TriggersSchedules and events

Run on a timer, when an email lands, when a record changes, or when a person asks.

DataContext, connected

Agents read your systems in place, with only the access each job needs.

ToolsConnectors and APIs

Email, CRM, ERP, files, data warehouses and your own APIs.

ModelsAny model, per task

Frontier, open-weight or your own. Switch without rebuilding the agent.

RuntimeBuilt for production

Sandboxes, memory, scheduling, retries, hosting and scaling.

Hand-offPeople where it matters

Your rules decide when an agent stops and a person decides.

Policy engineYour rules on every step

Limits, permissions and approvals checked before an agent acts.

ChecksWatch on every result

Every output graded from outside the agent, not by the agent itself.

RecordEvery step, kept

Who did what, on which data, approved by whom.

Any model

No lock-in. Ever.

We don’t sell the models, so your agents run on whichever works best.

Your agentModel AModel BModel C
Governance built in

Your policies, on every step.

Permissions, checks and a record of who approved what, on every run.

ReadDecideActRecordEVERY STEP PASSES YOUR POLICY
Next in the loop03Watch →Measure every result, graded from outside.
Early access

Build on it before everyone else.

We are opening Build to a small group of teams first. Tell us what you would build, and we will email you when your spot opens.

Book a demo

See the loop on your work.

Thirty minutes with Jack or Bill, our founders, on the workflow you would run first.