For operations teams
The process exists. It just isn’t written down anywhere.
You know the sequence: who to ask, in what order, what to check first, who to chase. That knowledge is real infrastructure — and right now it only runs when you’re at your desk. Requester turns it into something that runs without you.
Free plan, no card required. Submitters are always free.
What operations gets
- Canvas
- Nine node types, drag-and-drop, no scripting
- Rules
- Conditional branches on any field or metadata
- Time
- Durable delays, step SLAs, workspace escalation
- Vocabulary
- Custom statuses mapped to semantic categories
- Versioning
- Draft → publish, history, revert, clone
You are the workflow engine, and you don’t scale
Every process that depends on you noticing something is a process that stops when you're on leave, in a meeting, or looking at something else.
Processes that live in your head
- You're the routing table: people ask you who needs to approve this.
- Nothing is followed up unless you personally remember to follow it up.
- Two people run the same process differently, and both think theirs is right.
- Onboarding a new coordinator means transferring a year of undocumented judgement.
Processes that run themselves
- Routing is a workflow, applied identically whoever submits.
- Step SLAs remind, and escalation raises what the reminder didn't move.
- One published template means one version of the process, with history.
- A new coordinator reads the canvas instead of shadowing you for a month.
Four things that stop depending on you personally
Knowing who approves what
The branch reads the field and routes. Nobody has to ask you what the threshold is, ever again.
Remembering to chase
Durable step timers and workspace escalation handle the follow-up, on a clock rather than on your attention.
Keeping the process consistent
A published, versioned template is the process. Changing it is a deliberate act with a history and a revert.
Explaining where things are
The workflow map on each request shows the path it took and the step it's sitting on, without you narrating it.
Operations usually inherits all of them
These rarely have a natural owner, which is how they end up with you. The upside is that you're the one person positioned to make them consistent.
Vendor onboarding
Suppliers assemble their own file through a public link; legal and finance approve in sequence.
Learn morePurchase approvals
Threshold routing with the vendor and cost centre captured at source.
Learn moreClient & external intake
One public front door so outside requests stop arriving at whoever the client knows.
Learn moreRecurring internal requests
Equipment, facilities, and access asks, with request-again and smart pre-fill for the repeats.
Learn moreA process that only runs when you’re watching it isn’t a process. It’s a habit with a single point of failure.
How this actually gets adopted
Process projects fail on adoption, not on capability. The sequence below is the one that works, and it deliberately starts small.
Pick the one that annoys you
Not the biggest process — the most irritating one. Momentum matters more than coverage.
Scope it to one team
Templates activate per team. Prove it with a group that will tell you the truth.
Let people ask however they want
In-app, AI-drafted, email intake, or a public link. Adoption you have to enforce isn't adoption.
Duplicate what worked
Copy the template into the next team and adjust the approvers. The shape is already right.
What operations teams ask
- How long does it take to build a process?
- Minutes, not a project. Describe it in a sentence — “purchases over $500 go to my manager, over $2,000 also need finance” — and the AI generates the form, the branches, and the reminders. Then you edit whatever it got wrong on the canvas.
- Can we use our own status names?
- Yes. Custom statuses let you rename and recolour the defaults and add your own — “Pending Legal”, “PO Issued”, “Awaiting Delivery”. Each maps to one of nine semantic categories so the engine, SLAs, and analytics keep working underneath your vocabulary.
- What happens to a process when the person who owns it leaves?
- Nothing, which is the point. The process is a published template and workflow with version history, not knowledge in someone's head. Route to teams and reviewer groups rather than named individuals and the process survives any single departure.
- Do delays and reminders survive a deploy?
- Yes. Timers are durable and processed by cron rather than held in memory, so a three-day delay survives restarts and scaling events. Execution is also concurrency-safe — simultaneous decisions can't double-advance a request.
- Can we roll a process out to one team first?
- Yes. Templates can be scoped to specific teams rather than the whole workspace, and duplicated into another team once they work. That's the normal path: prove it with one group, then copy it.
Describe one process. See it running before lunch.
A sentence becomes a form, a routing rule, and a reminder schedule. Start with the process you're most tired of explaining.