Requirements, evidence and gaps

How can you link each compliance requirement to a source, evidence and a verifiable action?

1250 € per session, excl. VAT

In short

In a half-day session of 3 hours you turn a document file into a usable compliance table. Generative AI helps extract and rephrase requirements without determining their legal scope. You structure the evidence, document gaps and measure the quality of the deliverable before validation by a qualified professional.

Why this training?

A list of obligations alone is not enough to demonstrate compliance. Without a precise reference, an identified owner or expected evidence, teams do not know how to explain a status or prioritise outstanding checks. AI can suggest a useful extraction, but it can also merge two requirements or invent an interpretation. The training therefore establishes a line-by-line check and systematically distinguishes between missing evidence, an identified gap and uncertain applicability. Ongoing case study: The fictitious organisation « Néravel Pièces » is preparing a compliance review of its purchasing process. Its training file brings together simulated contractual clauses, an internal procedure and incomplete supporting documents. The group must build a matrix of ten requirements without confusing contractual commitments, internal rules and regulatory obligations that require confirmation. Application at work: Reuse the template within a limited professional scope with authorised, anonymised data. Have the sources and applicability validated by a qualified professional, then compare traceability and completeness rates before and after correction.

Objectives of this training

  • Define the documentary scope and exclusions of the review
  • Extract individual requirements with their source references
  • Structure the tracking, evidence and responsibility columns
  • Assess gaps without automatically concluding that they constitute non-compliance
  • Measure the traceability and completeness of the matrix

Who is it for and what are the prerequisites?

Roles and tasks concerned

  • Compliance officers and legal professionals involved in document reviews
  • Quality, internal control or purchasing managers responsible for gathering supporting documents

Prior skills

  • Have previous experience of making simple requests to generative AI
  • Be able to work with a table and distinguish between a procedure, a contract and a supporting document

What you will work on

  1. Define the scope of the document review

    Objective: Establish what the matrix must cover. Scope, document versions, source categories and exclusions; distinction between a requirement and a recommendation.

  2. Extract without inventing

    Objective: Obtain ten traceable individual requirements. Extraction instruction limited to the documents provided; quotation, location, rephrasing and flagging uncertainties.

  3. Link requirements and supporting documents

    Objective: Build a table that can be used during a review. Columns: identifier, source, requirement, applicability, check, expected evidence, available evidence, owner, status, action and deadline.

  4. Check and prepare for validation

    Objective: Measure defects and organise how they will be addressed. Checking references, duplicates, missing fields and status consistency; version control and responsibility for updates.

What you take away

  • Scope sheet and source register
  • Matrix of ten requirements with evidence, owners and statuses
  • Log of ambiguities and questions for legal validation
  • Quality control checklist and correction plan

Method and assessment

  • Brief instructional input followed by immediate practice using the ongoing case study.
  • Individual work followed by work in pairs, with participants reviewing each other’s outputs.
  • Guided group discussion: what works, what needs correcting and what remains a human decision.
  • Preparation: Arrange access to generative AI and spreadsheet software authorised by the organisation. Consult the fictitious file provided; any personal example must be anonymised and authorised for sharing.

Observable criteria: Each requirement has an identifier and a reference that can be found in the file. ; The ten rows distinguish between requirement, expected evidence, available evidence and status. ; Every uncertainty about applicability is flagged without an automatic legal conclusion. ; Each documented gap includes an action, a proposed owner and a deadline to be confirmed. ; The rates of sourced rows and completed mandatory fields are calculated, then defects are corrected..

Frequently asked questions

Does AI determine which obligations apply?

No. It helps structure the sources; applicability still needs to be checked with your legal department.

Do you need to use confidential documents?

No. The exercises use the fictitious file or anonymised documents within authorised tools.

Does missing evidence mean non-compliance?

Not necessarily: it primarily indicates a check that needs to be carried out and documented.

Does the matrix constitute legal advice?

No. It is a monitoring tool whose conclusions must be validated by a qualified professional.

Last updated: