Runbooks that keep up with reality
The SOP Maintainer keeps your runbooks current, flags where the documented process has drifted from what teams actually do, and marks a section for owner review when it cannot confirm the truth.
Tool changed · new ticketing macros
SOP MaintainerThe ticketing tool shipped new macros. Checking which runbook steps reference the old ones.
Reading the support escalation runbook3sReading the vendor’s change notes for the new flow3sSteps 3 and 4 name macros that no longer exist. The rest of the runbook still holds.
Rewriting steps 3 and 4 to the new macro names2sFlagging the update to the support lead to confirm1s
How it works
Tool changed · new macrosPolicy updated
Quarterly SOP reviewTeam: “this is out of date”2 steps updatedChange note datedDivergence flagged to ownerUnconfirmed marked for reviewTriggered by a change or a review cycle
A tool change, a policy update, or a scheduled review wakes the agent to check the runbook against how work is actually done.
Spots where the docs have drifted
It compares the documented steps to what teams really do, updates the approved way, and flags where practice has diverged from the SOP.
Writes back to Notion and Docs
Updated steps land in the runbook with a dated change note, and anything it cannot confirm is marked for owner review, not guessed.
Without Relevance
With Relevance
Runbooks are written once and rot until someone hits a step that no longer works.
The agent checks the runbook against reality on every relevant change and keeps it current.
Nobody notices the SOP and the real process have diverged until an incident exposes it.
The agent compares documented steps to how teams actually work and flags the drift.
When two teams do a step differently, the doc reflects whoever wrote it last.
The agent documents the approved method and flags the divergence to the process owner.
Only the runbooks someone champions stay current. The rest quietly go stale.
Every runbook gets the same review on every change, not just the popular ones.
Update quality depends on who last edited the doc and how careful they were.
One standard for structure, change notes and sourcing on every runbook.
A gap gets filled with a plausible-looking process nobody actually follows.
What cannot be confirmed is marked for owner review, never invented to look complete.
Held to a quality bar, on every run
Every runbook update is checked against eval test cases written for this exact job before it is published. If a run fails the bar, it never ships.
SOP MaintainerSampling 2% of live runs · 96% passing this week
Shaped by your team, not a template
Experts and engineers tune SOP Maintainer to your playbooks. Every change ships through the same evals.
Drag and drop
Compose agents and workforces on a visual canvas. Drag in agents, tools, and approvals in a flow anyone on the team can read.
Drag and drop
Compose agents and workforces on a visual canvas. Drag in agents, tools, and approvals in a flow anyone on the team can read.
Ticket TriagerBuild with AI
Describe the agent in plain language and Invent builds it: the prompt, the tools, and the evals to prove it works.
Build with AI
Describe the agent in plain language and Invent builds it: the prompt, the tools, and the evals to prove it works.
Claude Code
CodexBuild with MCP
Engineers drive the same platform from Claude Code, Codex, or Cursor. Create agents, link knowledge, and run evals over MCP.
Build with MCP
Engineers drive the same platform from Claude Code, Codex, or Cursor. Create agents, link knowledge, and run evals over MCP.
Built for enterprise teams
Run agents on your real data with the access controls, audit trails, and residency guarantees enterprises require all built in.
Security & data privacy
- Data residency
- PII masking
- Audit logs
- No training on your data
Access & controls
- Role-based access control
- SSO / SAML
- Human-in-the-loop approvals
- Version control
Monitoring & oversight
- Real-time monitoring
- Full agent tracing
- Cost visibility
- OTEL & Delta Share export
"The ability to be vendor agnostic and the ability to scale across a breadth of functions is a really key feature."
Levi Watters
Partner, KPMG Australia
"The key for us was how we can modularize industry knowledge and the best playbooks, and apply it."
Allen Roh
Senior Marketing Manager, Autodesk
"We're looking for every place where AI can allow sellers and customer success reps to be more engaged with customers."
Rob Giglio
Chief Customer Officer, Canva
"The ability to be vendor agnostic and the ability to scale across a breadth of functions is a really key feature."
Levi Watters
Partner, KPMG Australia
Team it up with more operations agents
New customer signed- 1Assigned six owners
- 2Nudged the late step
- Process on track
Process Orchestrator
Runs cross-team processes and chases every step.
Renewal · 60 days outVendor Renewal Tracker
Flags renewals before the deadline that binds you.
Salesforce
HubSpotNotionOne recordData Hygiene Steward
Reconciles records and flags what it cannot resolve.
Anomaly · backlog +38%Ops Reporting Analyst
Builds the weekly report and explains the anomalies.
Every agent you add shares one stack
These agents all run on one platform: one gateway, one router, one eval suite, one audit trail. Every agent after the first ships faster.
- Triggered by key events or signalslike Zapier
- Given access to contextlike Zep
- Connected to your appslike Composio
- Access to all LLMslike OpenRouter
- Performance evaluatedlike Braintrust
- No-code agent builderlike Dust.tt
- Coordinated into teamslike CrewAI
- Kept alive through long-running worklike Temporal
- Traced at every step of every runlike Langfuse
Put the SOP Maintainer to work this quarter
We partner with enterprise teams to take this work off their plate. An agent built with you, proven against your quality bar, and live in weeks.