Someone mentions an example in the meeting. A call center that deployed a chatbot, an insurer that assesses claims automatically. Everyone nods, and then there's a silence, because no one in the room knows what that means for their own department. Is the customer service work here comparable? Is the time-consuming part in looking up information, in typing responses, or in assessing a difficult customer? Without that breakdown, every conversation about AI remains a conversation about other companies. The work scan puts your own work under the knife of that same question, task by task, so the answer no longer applies to some other company.
Once the scan is complete, there is a report with a fixed structure. Not advice in running text, but an account you can page through and check at every point.
The task breakdown. Every function or department included is broken down into the tasks the work actually consists of in practice. Not the job title, but the activities: answering an email, checking a file, adjusting a schedule, drawing up a quote.
The assessment per task. Each task is accompanied by an estimate of the extent to which AI can now take over that task, partly support it, or cannot yet touch it. That estimate comes with an explanation: why this task does or does not qualify, and which assumptions were used.
The fte calculation. The outcomes per task are added up into a picture in fte, with a bandwidth instead of a single definitive number. The report shows how that bandwidth came about and which uncertainties it contains, so it is not a stray figure you have to believe on the tool's say-so.
The substantiation per outcome. For every result, you can trace back which data it was based on. No black box that spits out a percentage, but a trail back to the source. Exactly how that breakdown and assessment work is explained on the page how it works; the kind of results this produces in practice can be seen at outcomes.
The scan works with what already exists: job descriptions, process descriptions, system data on time spent, and where available, the data from the systems that record the work. No interviews, no sessions with managers to jointly determine what a task "actually" entails. The scan takes what is written in existing documents and datasets, places it side by side, and derives the outcome from that.
That is precisely why every outcome is traceable. A percentage or an fte estimate without a source is an opinion in the form of a number. This scan works the other way around: first the data, then the breakdown, only then the assessment. If you want to know why a task has been labeled "partly automatable," the report states what information that judgment rests on. More background on the way tasks are classified and weighted can be found in the knowledge base.
This is not tailored advice about what should happen in your company. The report contains no recommendation to reorganize, automate, or redeploy staff. The scan describes what the tasks are and what could potentially be done with them, not what should be done.
Nor is it a promise. Nowhere is it guaranteed that a task will actually be taken over by AI, or that a stated fte bandwidth will actually become available. The outcome is an estimate based on the current state of AI and the data supplied, not a commitment about the future.
It is not a conversation. No consultant comes along for a day, no one is interviewed, and nothing is "gathered" in a workshop. Everything in the report is machine work based on datasets. Human work only comes into play if you choose it yourself, and only after the report, with a partner you select yourself.
Finally, it is not an endpoint that forces you onto a particular route. The report is a dossier you can continue with yourself, at the pace and with the approach that suits your company.
With the report in hand, there is a choice, not an obligation. You can get to work yourself with the route map that provides the steps, templates, and time indications: what you can pick up yourself, in what order, and what each step requires in terms of data quality and integrations. You can choose to have the tool remain the dossier while a partner takes on the heavier steps, with the choice per component of whether you outsource it. And you can have a partner take on the entire implementation, with the report as a starting point instead of a blank slate, but only if you give permission for that yourself. In all three cases, it begins with the same document: the breakdown of your work into tasks, with the assessment and the calculation that follows from it.
Vraag maar. Ik ken de kennisbank van deze site; wat ik niet weet, zeg ik erbij.
Answers come from this site’s knowledge base. Not tailored advice, and not a scan of your company.