Practical AI guides

Decide, verify and automate with a method.

38 independent guides for turning an artificial-intelligence promise into a testable decision. Each method connects a precise intent, practical criteria and tool profiles linked to its subject.

Visual paths for choosing, evaluating and governing artificial intelligence tools.
AI-generated illustration
38
in-depth guides
212
tool profiles
6
thematic paths

Choose a path

Pick a path for the decision at hand. Each guide also stands alone and can be read in another order.

Bookshelves and hanging lamps in a library in Paris.
Real photograph: Janko Ferlič, library in Paris. Source: Unsplash · Unsplash License
AI-generated illustration
Governance guide · 12 min

Build useful AI governance for a team

A useful policy does more than approve or ban a tool. It connects real uses to risk, an accountable owner, expected evidence and a clear response when something fails.

Read the guide
AI-generated illustration
Accessibility guide · 11 min

Design accessible and inclusive AI experiences

An AI-assisted feature remains inaccessible when its interface, answers or media exclude part of the audience. Accessibility must be checked across the full journey, from entering an instruction to correcting the output.

Read the guide
AI-generated illustration
Vendor guide · 12 min

Select and qualify an AI solution provider

A convincing demonstration proves neither production quality nor provider resilience. Qualification must address a specific offer, defined use, dated evidence and realistic exit conditions.

Read the guide
AI-generated illustration
Adoption guide · 10 min

Train a team and build useful AI adoption

Distributing licences creates neither competence nor value. Adoption grows when each role can recognise a suitable task, use an approved workflow, verify the result and ask for help without hiding mistakes.

Read the guide
AI-generated illustration
Monitoring guide · 12 min

Monitor AI quality and respond to incidents

An AI workflow can keep responding while its quality deteriorates. Monitoring must track business quality, errors and consequences, then support shutdown, diagnosis and recovery without improvisation.

Read the guide
AI-generated illustration
Verification guide · 9 min

Fact-check an AI answer before reusing it

A persuasive answer can combine accurate facts, irrelevant sources and overconfident conclusions. Verification starts by isolating important claims and tracing them to original documents.

Read the guide
AI-generated illustration
Customer support guide · 11 min

Put AI to work in customer support

A quick reply does not necessarily solve a customer’s problem. AI-assisted support needs current documents, a clear way to acknowledge limits and a smooth handoff for difficult cases.

Read the guide
AI-generated illustration
Practical method

Validate AI structured outputs before importing data

Well-formed JSON can attach a plausible date to the wrong document. When extracting text into a spreadsheet, form or database, check the API response, the JSON structure and the relationship between each value and its source separately.

Read the guide
AI-generated illustration
Practical method

Check commitments in an AI-generated meeting summary

An accurate transcript can still produce poor minutes if it turns an idea into a decision. Fathom and tl;dv document summaries and retrievable passages; validating commitments requires reading context rather than searching only for action items.

Read the guide
AI-generated illustration
Practical method

Check document permissions in a connected AI assistant

Even when the original application enforces file permissions, check what the connected assistant exposes in citations, previews and cached answers. Before connecting team documents, compare each role’s actual access in the source with the information that role receives from the assistant.

Read the guide