Resources

A little method. Much more clarity.

Choose a question you are facing today.

01

Your first AI project: a 90-day roadmap

A useful first project starts with a recurring task and a clear decision. This 90-day outline helps organise the work. Adjust its duration to your data, approval requirements and team availability.

Explore ↗
02

Questions to ask before choosing an AI tool

A feature list is not enough to compare two offers. Frame your questions around the work, the data entrusted to the supplier and daily operation. Ask for written answers so commitments can be checked later.

Explore ↗
03

Measuring the value of an AI project: a calculation framework

An estimate supports a decision when observations are separated from assumptions. Use this framework to prepare a pilot, then check whether it remains worthwhile after human review, exceptions and recurring costs are included.

Explore ↗
04

Helping teams adopt AI

Adoption starts with everyday work. Useful training shows when to use the tool, how to check its proposal and what to do when it is unsuitable. Begin with one team and situations its members recognise.

Explore ↗
05

A 75-minute workshop to identify AI opportunities

Bring together an owner and the people doing the work. This internal workshop selects one opportunity to investigate without choosing a solution too early. Prepare a shared board and a few non-confidential examples.

Explore ↗
06

Your AI pilot has stalled: where is the blockage?

A pilot can produce a convincing demonstration and still be difficult to use. Before adding a tool or changing models, examine the problem, access to information, fit with the workflow and the decision expected.

Explore ↗
07

Existing feature, off-the-shelf product or custom integration?

These three options suit different situations. Compare them against the same need, test set and cost horizon. The choice should make sense to both the business team and the person responsible for operation.

Explore ↗
08

Confidentiality, access and approval: preparing controlled use

Before connecting an assistant to documents, map the information flow and the decisions it may influence. This guide helps prepare discussions with your business, security and privacy owners. Requirements specific to your activity still need to be examined.

Explore ↗
09

Keeping control: designing an AI system you can leave

The ability to change solutions starts before deployment. Reversibility covers data, rules, connections and the skills needed to keep working. An available export alone does not guarantee a usable restart.

Explore ↗
10

AI watch 2026: turn an announcement into a useful decision

References checked on 12 October 2026. A development deserves attention when it changes a need, cost, risk or way of working. This guide provides a monitoring framework for the tools your business actually uses.

Explore ↗

One process. One next step.

Start with your daily work to decide what is worth checking together.