One person is enough
The task goes to every recipient and the first answer closes the step. The other tasks are cancelled.
A refund to approve, access to authorise, a case involving several departments. Map the process and track tasks, ownership and progress directly in Uppo.
Examples
A refund to approve, access to authorize, a case involving several departments: each follows a traceable process.
Workflow builder
Combine tasks, decisions, waits and connections to business systems. Before publication, Flow validates the structure and flags issues to address.
The starting point of the procedure. One per Flow.
Collect information through dedicated fields or an existing ticket form.
Request a decision from the owners and define what happens if it is rejected.
Choose a path using ticket data and values collected in earlier steps.
Assign, update, notify or call a webhook without manual intervention.
Pause for a duration or until a date, then resume automatically.
Start parallel activities when several teams can work at the same time.
Bring parallel paths together before the next activities.
Open a ticket for another team while keeping it linked to the case.
Pause until a linked ticket reaches the specified status.
Send a request to an external system and handle success and failure separately.
Close with a declared outcome and optionally update the linked ticket.
Essential workflow components are included on every plan. Advanced nodes are available with Enterprise.
Approvals
Choose who decides, whether one response is enough or several approvals are required, and what happens after a rejection.
The task goes to every recipient and the first answer closes the step. The other tasks are cancelled.
The step closes once every recipient has answered. Use it where approval is collective.
The task reaches one person at a time, in the order you listed the recipients. The next one gets it after the previous answer.
Each task has a due date. Once it passes, the task stays marked as overdue in its owner’s task list.
If the person expected to respond has an active delegation, the task reaches the delegate without redrawing the Flow.
Each person has a list of steps waiting for them, with the related case and due date.
A refund to approve, access to authorise, a case involving several departments. Map the process and track tasks, ownership and progress directly in Uppo.
Assign work, collect information and manage approvals while keeping the context intact between departments.
Define the steps and rules once. Your team follows the same process for every request instead of rebuilding it case by case.
Assign tasks to people or teams and set deadlines. Ownership and the next action are always visible.
See progress, open tasks, decisions and history without asking the team for updates.
Assign a form to the right person and let the workflow continue only when the required data is complete.
Define the fields on the step itself, or reuse a ticket form you already built and avoid writing the same form in two places.
The step goes to a specific person or to a team. If you name several recipients, the engine opens a task for each one under the rule you chose.
Every task carries its due date. Once it passes, the task stays flagged as overdue in the lists of whoever owns it and of the manager.
What someone writes in a form becomes a value on the case, readable by the conditions and actions that come after.
If the person who should answer has an active delegation, the task reaches the delegate without anyone redrawing the Flow.
Each person has the list of steps waiting on them, with the case they came from and the due date.
Conditions send each case down the right path. Independent tasks run together and rejoin when they are complete.
Compares ticket fields and the values collected in earlier steps. Rules combine by requiring all of them or accepting any one, and a branch can be the default for cases that match none.
A branch node opens two or more routes that run at the same time. This is the case of two teams doing their part without waiting on each other.
The path resumes once every branch opened by that split has reached the join, and continues as one until the outcome.
The validator confirms that every branch returns to its join and that none of them ends on its own: a case never waits for something that will not arrive.
Combine assigned tasks and approvals with automatic ticket updates, notifications, emails and webhooks.
This is the default: the error is recorded on the case and the path continues. Right for a notice or a note.
Switch it on for a single node. The case stops and is flagged, instead of assuming an action the following steps depend on, such as a refund.
A wait node pauses the path for a duration (24 hours, 2 days) or until a specific date. Then the engine resumes on its own.
Every step opened, every answer, every action and every error goes into an append-only log you can read back to reconstruct how the outcome was reached.
Tasks past their due date stay flagged as overdue, and stalled cases have a view of their own.
On closing, the outcome node can leave a note on the linked ticket or move it to a status you choose.
An agent can launch it from a ticket, or a rule can start it automatically when the expected conditions are met.
An agent starts the procedure from a ticket when they recognise the case. It takes the permission to start Flows: seeing them and starting them are two separate rights.
A rule starts the Flow when its conditions are true. This is the route for procedures that must always run, without depending on who opens the queue.
The rule covers expected cases; agents can start the procedure manually for other cases.
On the same ticket you decide whether one case at a time is enough or several can coexist. Procedures can also be started from the API.
Work in draft, fix configuration errors and publish only when the path is complete and valid.
Design and save as often as you like. The draft runs nothing and touches no case in progress.
The validator reviews nodes, connections, branches and settings, and points you to the node behind every error or warning.
The draft becomes the published version. The check runs again on exactly what you are publishing.
Every case freezes the version it started with and follows it to the end. A new version applies to the cases created after it: nobody halfway through a path finds different steps in front of them.
A Flow can be deactivated without deleting it: it stops being startable and stays available to reread and switch back on. Cases already open carry on.
Monitor tasks, deadlines, decisions and overall status from one view.
Each person sees the steps waiting on their answer, with the due date and the case they came from.
Every case with its status: running, waiting, blocked, completed, cancelled or failed.
A dedicated view for the cases that are stuck, with the reason for the block.
Completed steps and open ones, the values collected, the branches running and the full history.
A case started by mistake can be cancelled, under a permission separate from the one to start it.
On closing, a case carries a stated result — approved, declined, or whatever label makes sense for your procedure.
Use them together: a rule recognises the case and starts the workflow; Flow coordinates people, deadlines and approvals through completion.
Rules that react to an event on a ticket. They check conditions and take one or more actions, straight away. Nobody has to step in: the ticket is updated and the rule is done.
A procedure that runs over time. It has steps assigned to people, approvals, conditions, waits and a final outcome. It stays open until the path reaches the end, and at any moment you know where it is.
An automation starts a Flow when a condition is met, and an action node inside the Flow uses the same catalogue of actions as the rules. They are two ways of triggering the same work.
See how automations workBuild workflows with forms, approvals and linked tickets. With Enterprise, you can also call external APIs and transform data between steps.
The core canvas nodes are part of Starter, Growth, Scale and Enterprise. The canvas is not reserved for the higher plans.
Start from a template or design the process on the canvas. Steps, owners, deadlines and approvals remain editable by your team.
HTTP calls to external systems and data transformations are reserved for Enterprise. Linked tickets, waits and human steps remain in the core catalogue.
Setting up, starting and managing procedures.
Automations apply actions when an event occurs and the configured conditions are met. Flow manages a procedure over time, with assigned tasks, approvals, waits and an outcome. An automation can start a Flow.
It depends on how you configure it. A Flow can be startable only by hand, only by an automation, or both ways. Manual start requires a dedicated permission, separate from seeing procedures and from designing them. Procedures can also be started from the API.
To a person or to a team. If you name several recipients, the engine opens a task for each under the rule you chose on the node, and if the person who should answer has an active delegation the task reaches the delegate.
On each approval you choose whether one person is enough, whether everyone must answer, or whether they answer in order. You can set a due date: once it passes the step shows as overdue, and the decision still belongs to a person. On a refusal, you decide whether the procedure ends, goes back to the start, returns to the previous step, or jumps to a specific node.
Yes. You can open parallel work paths and bring them together before continuing. Publication checks verify that the paths reach the intended join point.
Configure the behaviour on each node. A non-blocking action records the error and lets the path continue; a blocking action stops the case and reports the issue. Choose according to how the action affects later steps.
Each case retains the published version it started with. New versions apply to cases started after publication.
Start from a template or design the procedure on the canvas, defining steps, owners, deadlines and approvals. Core nodes are included on every plan; advanced nodes remain reserved for Enterprise.
Core Flow is included on every plan, from Starter to Enterprise. Advanced HTTP request and data transform nodes are reserved for Enterprise.
The rules that react to an event on a ticket, and that can start a Flow when a procedure is needed.
See the automationsChannels, tickets, SLAs, knowledge base, portals and integrations: the whole product at a glance.
See the featuresWhat Uppo costs, what each plan includes and how the agents in the plan work.
See the pricesStart with an example, adapt it to your service and publish it when it’s ready.