Example cases: what changes when the bottleneck is gone
These six cases are anonymized, realistic scenarios for Dutch companies with 10 to 100 employees. The companies are described, not named. The figures are illustrative and fall within what we see during scans and prototypes. Every case follows the same route: situation, bottleneck, what we built, and the result in hours.
Why we use example cases
We don't name clients without permission, and we don't round figures up to impress. That is why every case carries the same label: Example case · anonymized · figures are illustrative. What is real: the processes, the systems, the exceptions, and the order of the work.
EXAMPLE CASE · ANONYMIZED · FIGURES ILLUSTRATIVE
Every case follows the same route: situation, bottleneck, what we built, and hours saved.
Filter by sector
6 cases
Technical wholesaler (45 employees): 33 hours a week returned to sales support
SITUATION
About 180 orders a day, 60% of them as an email with a PDF or Excel attachment. Sales support retyped every line into Exact, with peaks on Monday.
BOTTLENECK
42 hours a week of retyping, errors in 2.6% of order lines, and orders waiting 2 to 4 hours for a confirmation.
SOLUTION
An inbox listener reads orders from PDF, Excel, and email text, matches the client and item numbers in Exact, and sets them up on a review screen.
RESULT
42 → 9
sales support hours/week on order entry
+3 more
Hospitality wholesaler (28 employees): price lists in 18 hours instead of 120
SITUATION
Forty suppliers send price lists in Excel and PDF every quarter, each laid out differently. One buyer updated 12,000 items by hand.
BOTTLENECK
About 120 hours a quarter on price lists, pricing errors discovered by clients, and 25 hand-written backorder emails a day.
SOLUTION
A list standardizer that recognizes columns and flags changes above 15%, a review screen, and a bulk update to the ERP.
RESULT
120 → 18
hours/quarter processing price lists
+3 more
Transport company (60 employees): invoice 3 days after delivery instead of 12
SITUATION
About 350 trips a week. Drivers photograph the CMR or POD; administration looks up the matching trip in the TMS for each photo and invoices at the end of the month.
BOTTLENECK
22 hours a week matching and chasing missing PODs, invoices on average 12 days after delivery, and 6% of invoices disputed for lack of proof.
SOLUTION
A photo and PDF intake, AI reads the CMR fields and links them to the trip automatically. Missing PODs are on a list for the planners the next morning.
RESULT
22 → 6
administration hours/week on CMR processing
+3 more
Logistics provider (35 employees): planners spend 17 fewer hours a week on email
SITUATION
Transport requests come in as free-text emails and through three client portals. Planners retype them into the TMS and answer "where is my shipment?" questions all day long.
BOTTLENECK
15 hours a week retyping requests, 40 status questions a day (9 hours a week), and planners interrupted every 10 minutes.
SOLUTION
A shared planning inbox from which an AI step pulls the request fields and sets up a prefilled TMS order.
RESULT
24 → 7
planner hours/week on email
+3 more
Accounting firm (18 employees): 91% of clients complete on time
SITUATION
Every quarter, employees chase bank statements, invoices, and payroll documents by email and phone. Documents come in through email, WhatsApp, and a portal.
BOTTLENECK
12 hours a week calling and emailing, 30% of clients late, a peak in the final week, and a tracker that is never right.
SOLUTION
Automatic reminders based on the open items in Exact Online and Twinfield, an intake inbox that recognizes attachments and files them by client, and a small app as the tracker, with escalation to the account manager.
RESULT
12 → 3
hours/week collecting documents
+3 more
Accounting firm (55 employees): one secure business AI instead of 55 personal accounts
SITUATION
Employees used personal ChatGPT accounts for draft letters and tax questions. Management was worried about client data and the EU AI Act.
BOTTLENECK
No policy, no logging, no source references. More than 20 minutes per routine client question, and no way to tell whether 55 people met the AI literacy duty.
SOLUTION
A business AI workspace on EU-hosted models, connected read-only to SharePoint and the practice-management system, roles per client group, answers with a source.
RESULT
20 → 6 min
per routine client question
+3 more
