Approvals without bottlenecks

How can you reduce waiting times and requests for rework in a purchasing approval process without weakening controls?

1960 € per session, excl. VAT

In short

This 7-hour day, excluding breaks, enables you to review a purchasing approval process with support from generative AI. You map the steps, identify bottlenecks, formalise decision rules and test a proposed process using fictitious case files, with indicators for turnaround time and quality.

Why this training?

An approval process can accumulate approvals without clarifying who checks the budget, supplier compliance or operational need. Incomplete case files then circulate between departments, and urgent requests sometimes bypass the rules. Generative AI helps compare process routes and make inconsistencies explicit, but its proposals do not constitute authorisation. Every simplification must remain compatible with delegated authority, segregation of duties and the traceability of decisions. Running case study: The fictitious organisation « Asterive Outillage » purchases maintenance parts. Its requests go through the workshop manager, purchasing and finance. Twelve fictitious timestamped case files reveal missing documents, redundant approvals and bottlenecks during absences. A supplied policy on financial commitments sets out the mandatory checks. Application at work: On your return, select a limited process, have purchasing and finance review the proposed rules, then compare turnaround times and returned case files over a defined period. Do not change any delegated authority without the agreement of the authorised manager.

Objectives of this training

  • Map the steps, stakeholders and information in an existing process.
  • Measure waiting times and returned case files using timestamped events.
  • Distinguish mandatory checks from redundant approvals.
  • Formalise thresholds, delegated authority and exception rules.
  • Verify a proposed process and document its indicators.

Who is it for and what are the prerequisites?

Roles and tasks concerned

  • Purchasing and supply managers dealing with stalled requests.
  • Operations managers and process leads responsible for approval processes.

Prior skills

  • Understand a purchase requisition process and its responsibilities.
  • Be able to query a generative AI assistant and verify its response.
  • Be able to calculate durations and filter data in a spreadsheet.

What you will work on

  1. Define the scope of the process under review

    Objective: Define the start, end and checks within the process. Admissible request, financial commitment, stakeholders, supporting documents and confidentiality.

  2. Establish objective evidence of bottlenecks

    Objective: Separate processing time from waiting time. Timestamps, returns for additional information, duplicates and limitations of the available data.

  3. Clarify decision-making authority

    Objective: Assign each check to the appropriate role. Financial commitment thresholds, segregation of duties, cover arrangements and escalation.

  4. Simplify process routes

    Objective: Reduce back-and-forth exchanges without removing safeguards. Completeness at submission, compatible parallel checks, reasoned rejection and exceptions.

  5. Test the selected rules

    Objective: Identify ambiguities before making any actual changes. Standard case files, exceeded threshold, urgency, absence and missing supporting document.

  6. Measure and hand over

    Objective: Enable those responsible to monitor the revised process. Median turnaround time, return rate, initial completeness and traceability of decisions.

What you take away

  • Annotated process map and analysis of waiting times.
  • Matrix of approvals, delegated authority and exceptions.
  • Proposed procedure with a list of required documents.
  • Verification checklist and indicators table.

Method and assessment

  • Short input sessions followed by immediate practice using the running case study.
  • Individual work followed by work in pairs, with peer review of the outputs.
  • Facilitated group review: what works, what needs correcting and what remains a human decision.
  • Preparation: Identify a process and compile an anonymised description of its rules, without sensitive supplier data. Check access to an AI tool authorised by the organisation; the exercises use fictitious or anonymised cases.

Observable criteria: Each step has an explicit owner, input and decision. ; Durations are recalculated from the supplied timestamps. ; No mandatory check is removed without justification and an identified decision on the trade-off. ; Incomplete and urgent case files, and those affected by an absence, follow an explicit rule. ; Each indicator specifies its formula, source and owner..

Frequently asked questions

Can AI approve a request?

Not in this approach: AI supports the analysis, while the decision remains with authorised people.

Do you need to change purchasing software?

No. The rules and supporting materials produced remain independent of any software change.

Can you work without usable historical data?

Yes, using the fictitious case files; actual gains cannot be established without a baseline measurement.

How should you handle an urgent request?

Provide an explicit route with the appropriate authority, justification and traceability, rather than an informal workaround.

Last updated: