The Four Constraint Families
Every job assignment in Checkfirst is evaluated across four families of constraint. These are not optional — all four are assessed on every proposed assignment before it surfaces to the scheduler.
Certification constraints are the hardest gate. A job requiring ISO 9001 lead auditor status will only surface candidates who hold that credential, with a validity check against the job date. Where a job requires compound credentials — for example, an IATF 16949 audit that also requires a client-specific witness assessment sign-off — all compound conditions must be satisfied simultaneously. Partial matches are excluded from the candidate list entirely, not ranked lower.
Geographic constraints operate as a combination of hard and soft rules. Hard rules define maximum acceptable travel distance or time for each job type (typically configured per scheme or per client contract). Soft rules express preference within that boundary — Checkfirst scores proximity relative to other candidates in the same region so that tightly-clustered assignments favour the geographically-closest available inspector. Where multiple jobs can be grouped into a logical trip for the same inspector, the engine proposes the clustering and calculates the consolidated travel estimate.
Impartiality constraints are enforced at the certification-body level as hard exclusions. When an auditor is flagged on the impartiality register for a specific client — whether due to prior employment, consulting relationship, or committee-defined conflict — that auditor will not appear in the candidate list for any job at that client site, regardless of certification or availability match. These exclusions are maintained per-client-per-auditor and are applied automatically without the scheduler needing to remember or look up the register.
Availability constraints encompass confirmed blackout dates, approved leave, daily hour caps, weekly overtime limits, and already-confirmed job commitments. The engine treats availability as a composite time-fence: an auditor with three committed audit days that week and a maximum of four chargeable days available will only be proposed for assignments that fit within the remaining window — and will show as unavailable if the remaining capacity is too short for the job duration, even if nominally available on the date in question.
How the Algorithm Scores Candidates
Within the pool of candidates that pass all hard constraints, Checkfirst produces a ranked match score for each. The score is a weighted composite of five factors:
Certification fit depth — a candidate holding the exact primary credential plus relevant secondary credentials scores higher than a candidate holding only the minimum required credential. This rewards assigning auditors with the most appropriate qualification portfolio, not merely the minimum valid one.
Geographic proximity — calculated as estimated point-to-point travel time from the inspector's registered location to the job site. Weighted relative to other candidates in the eligible pool for that specific assignment.
Availability margin — candidates with comfortable remaining capacity score higher than candidates at the edge of their weekly limit. This distributes workload more evenly across the pool and reduces the risk that a proposed assignment will need to be retracted due to a subsequent availability change.
Historical performance — where an auditor has a confirmed track record at a specific client or scheme, that history factors positively into their score. Conversely, documented client preference exclusions are applied as negative weight without removing the candidate entirely (unless combined with an impartiality hard exclusion).
Travel cost estimate — the predicted per-diem travel cost for the assignment, expressed as a normalised cost-efficiency rating relative to other candidates. This factor is weighted according to the travel cost cap configured for the job type.
What the Scheduler Sees
The scheduling interface presents a weekly assignment board with a prioritised candidate list for each open job. Each candidate entry shows their match score, the component breakdown (why this score), their current week's committed days, and the estimated travel cost for this assignment. The scheduler does not need to consult external systems — all information needed to confirm or adjust the assignment is present in the panel.
For bulk scheduling — for example, filling all jobs for a given week across a 40-person auditor pool — the scheduler can trigger an auto-fill run that proposes assignments across all open jobs simultaneously. The proposed schedule is presented for review before any confirmations are sent. Changes to individual assignments in the proposed draft recalculate the remainder of the week automatically.
Before and After: A Workflow Comparison
In a manual scheduling workflow, a typical week at a 40-auditor certification body follows a pattern that most ops managers recognise: Monday morning, the coordinator opens the master Excel, cross-references the new job list against the certification tab, manually checks the availability column, opens a separate spreadsheet for the impartiality register, drafts the week's assignments, shares by email for approval, makes revisions after three or four replies, sends confirmations, and updates the master file. That process — repeated every week — typically consumes 6 to 12 hours depending on volume and last-minute changes.
With Checkfirst, the coordinator opens the scheduling board on Monday morning and finds the week's jobs already queued with candidate rankings populated. They review the proposed assignments, adjust any that require judgment calls (unusual client relationships, split-trip logic, inspector workload preferences), and confirm. Notifications go to inspectors automatically. The board updates in real time. The same task takes 20 to 40 minutes. The difference is not that the coordinator is faster — it is that the constraint-checking work that consumed most of their time has been done by the engine before they open the screen.