Source documents and prepared records inspected in an operations workshop

RESULTS & REAL CASES / KALIITS

Less typing. Better follow-up. Real outcomes.

98% less time on WooCommerce listings, 80% less manual data-entry time, and every lead answered: explore Mehdi Kourchal’s projects and the scope of each outcome.

Discuss a controlled pilot

Reported project outcome

SLS: 80% less manual data-entry time

Illustration of the described workflow
  1. Email / WhatsApp
  2. OCR + review
  3. WMS

At the SLS warehouse, OCR and automation reduce time spent on manual data entry by 80%, according to Mehdi Kourchal’s field account. Data from inboxes, WhatsApp documents and digital files is inserted as it arrives, instead of piling up for manual batch entry into the WMS. One person checks cases requiring review; those exceptions are infrequent in the described workflow. Clerks can take on other warehouse tasks.

Source: Mehdi Kourchal’s field account. Scope and limits are stated in this account.

The experience behind Kaliits.

These cases come from work carried out by Mehdi Kourchal, before or through Kaliits. The previously reported outcomes were recorded on 2 October 2026. The real Autohaus Frisch / Michelin case background was confirmed on 4 October 2026. Each account keeps its own scope.

DEMONSTRATION • SYNTHETIC DATA

10 invoiced. 7 accepted. A decision to make.

This calculation runs the SIEMA application’s comparison rules. It uses no customer documents or connected systems.

Ordered
10 units
Delivered
8 units
Accepted
7 units
Invoiced
10 units
Damaged
1 unit
Invoice held for review

3 units invoiced beyond the accepted quantity. Short delivery and damaged goods require purchasing to review the evidence.

Calculated result: REVIEW_REQUIRED · 3 flags

How the decision is calculated

Accepted quantities below the order flag a short delivery. Any damaged goods require review. An invoice above the accepted quantity flags overbilling. Delivered quantities must equal accepted plus damaged quantities. All evidence must be present and consistent before the case moves to finance review.

Autohaus Frisch / Michelin: a real case, 20 ordered and one exception

Kaliits worked on a real Autohaus Frisch case involving Michelin: 20 items were ordered, with one damaged or missing. This is a real case background, confirmed by Mehdi Kourchal on 4 October 2026, rather than an invented purchasing scenario. This case concerns a receiving exception. No price, measured processing time or financial resolution is added to the account.

Try the receiving case above

Two synthetic cases execute the same comparison rules as SIEMA. Inspect the quantities, flagged differences and proposed next decision; payment remains subject to human approval.

Examples to examine

Each example needs representative cases, a named decision owner and acceptance criteria before use.

Import dossier readiness

Identify missing or conflicting pieces before a reviewer decides what to request next.

Explore

Purchasing mismatch

Compare ordered, received, accepted and invoiced quantities, then route discrepancies.

Explore

SIEMA receiving scenario

A separate illustrative purchasing and receiving application. It is not a customer result.

Explore

Document extraction

Extract invoice and shipment fields while leaving uncertain data for review.

Explore

Incident routing

Structure a field report, ownership and escalation path.

Explore

Lead and request intake

Capture an incoming request, flag missing information and assign the next action.

Explore

THE PROCESS AT A GLANCE

Evidence before a decision

Explanatory example; no customer outcome is implied.

Evidence objects presented for a named human review.
  1. Collect the source records
  2. Show the mismatch or uncertain field
  3. Record a human review and its limit
Project outcomes and interactive demonstrations.

Smatch, Wawa Creative, SLS and SRM IMS outcomes are reported by the systems’ developer. Tedi Services / Delassus is a real-case test. The SIEMA scenarios and diagrams above are separate illustrative examples. For your project, we define the baseline, processing time, follow-up coverage and required corrections for your own workflow.

Test one of these patterns with your own cases

Agree on access, exceptions, human approvals and a baseline before building.

Request a workflow diagnostic