Skip to content

Handover and exit

You can stop after any stage, and what was built for you is yours. This page covers what you own, what we hand over, how an in-house team or another vendor takes over and how your data is returned.

You own the custom code and the data built for you. In practice that means:

  • The software in the operations layer: the screens, queues, rules and integrations built around your workflows.
  • What was built for each agent: its instructions, its limits and the thresholds set with its workflow owner.
  • The data that software holds: the records, the corrections, the exceptions and the audit trail.
  • The four Discovery Sprint deliverables.
  • The ERP-like core, if we built one because there was no usable ERP.

We keep our reusable tooling: the general-purpose tools our forward-deployed engineers carry from one engagement to the next. Data ownership and portability covers where that line falls.

Your ERP and the data in it were yours all along. The ERP stays the system of record throughout.

Item What it contains
Source code All the custom code with its change history, in a repository you control
System documentation How the three layers fit together at your site. For each connection to the ERP: what it reads, what it writes and through which interface.
Runbooks Step-by-step instructions for the jobs that keep the software running: release a change, roll one back, restart a failed exchange with the ERP, use the stop switch, move a workflow down a level, change an agent’s limits
Workflow records For each workflow: its workflow owner, its current level, its limits and thresholds, and the record of every promotion review
Access inventory Every account and service account the software and the agents use, and every account of ours that has to be removed
Automation Backlog The current list: what is in use, what is in build and what has not been started

The documentation and runbooks are written as each workflow goes into use, then completed in the final weeks of the build. They are not written from memory in the last week.

If you stop after What you keep What happens
Discovery Sprint Operations Map, Systems and Data Audit, Automation Backlog, Working Prototype Our read access is removed. The deliverables are yours to use with your own team or with anyone else.
Embedded Build All of that, plus the custom code, its data, the documentation and the runbooks The takeover described below happens in the final weeks of the build.
Operate & Automate The same, as it stands on the day we stop The takeover described below happens before we stop running and monitoring.

The team that takes over can be your own or another vendor’s. The steps are the same.

  1. You name the team. They get the repository, the documentation and the runbooks.
  2. They shadow our engineers through a normal week: a planning session with a workflow owner, a release, the monitoring checks, a support problem.
  3. They run a release themselves while our engineers watch.
  4. Monitoring and support move to them. Alerts go to their people. Who held the AI model provider account and the hosting account during the engagement was agreed in your contract. If either was ours, it moves to you.
  5. Our access is removed. Access and credentials covers how.

Your workflow owners do not change. They still decide how each workflow should work and sign off changes to it, and promotion reviews carry on with the new team. A workflow owner who wants more caution while the new team settles in can move a workflow down a level on the day of the handover, then bring it back up through a promotion review.

Four kinds of data are involved.

  • Data in your ERP. The ERP stays the system of record, so there is nothing to return. We read from it and write to it only through interfaces it supports.
  • Data in the operations layer. This is yours: orders in queues, the documents that came in, corrections, exceptions and the audit trail. You keep the running software, a full export or both. Where that software includes parts of our reusable tooling, your contract gives you the right to keep running and changing that copy after we leave, with no license fee. Data ownership and portability lists the export formats.
  • Working copies our engineers made. Notes, photos, sample documents and data extracts used during the work are deleted. After exit we keep none of your data.
  • Data sent to an AI model provider. What the provider keeps, and for how long, is set by the provider terms reviewed with you: before day one of the Discovery Sprint for the Working Prototype, and before each workflow went live. AI providers and your data covers that review.

Where the software and its data are hosted is agreed in your contract. Stages and terms lists what your contract records. Whatever was agreed, a handover ends with your administrators in control and our access removed.