{"resourceId":"npcc-innovation-awards-west-yorkshire-post-call-ai-cleveland-redaction-20261021","versions":[{"version":"external-9f8d461cee7c14f1278f3c53a7ddc3e6c823f7177fbf86f41effb5e6a821775c","resource":{"id":"npcc-innovation-awards-west-yorkshire-post-call-ai-cleveland-redaction-20261021","title":"NPCC innovation awards report West Yorkshire Police’s AI post-call analysis at 97.6% accuracy versus 82% for human reviewers, and Cleveland Police’s AI redaction pilot cutting processing time 85–90%","organization":"National Police Chiefs’ Council and Police Digital Service; West Yorkshire Police; Cleveland Police","sector":"Policing / control rooms and digital evidence","geography":"England and Wales, United Kingdom","publishedAt":"October 21, 2026 (NPCC news release)","publicationDate":"2026-10-21","eventDate":null,"sourceName":"National Police Chiefs’ Council","sourceLabel":"National awards shine spotlight on the innovation transforming policing","sourceUrl":"https://news.npcc.police.uk/releases/national-awards-shine-spotlight-on-the-innovation-transforming-policing","evidenceClass":"public-sector-association","outcomeClass":"effective","topics":["knowledge-work","operating-model","accessibility-workforce"],"finding":"The NPCC release on its Innovation and Digital Summit awards says Post-Call Analysis, built by West Yorkshire and Hertfordshire police, the NPCC Digital Public Contact Programme and Deloitte, turns control-room audio into transcripts and summaries and flags vulnerability. Evaluation found 97.6% accuracy on key call characteristics versus 82% for human reviewers, vulnerability indicators identified in 99.3% of relevant cases, about 5.7 times more context captured and about 150 times more calls processed per hour; West Yorkshire deployed it operationally. Cleveland Police’s Pimloc Redact pilot cut per-file processing by about 85–90%, with human quality review. Warwickshire reported 42% of staff using Copilot Chat within eight weeks.","sledRelevance":"These are among the more specific measured results for police AI in contact centres and evidence handling, though reported in an awards release rather than a published evaluation. They give U.S. dispatch-adjacent police call review and records units concrete metrics to ask vendors for, and show human review retained in both deployments.","evidence":"Opened the NPCC release dated Oct. 21 and read all award entries, including the PCA accuracy and throughput figures, the Cleveland redaction pilot metrics and user ratings, Thames Valley/Hampshire automation savings and Warwickshire Copilot adoption figures.","architectureImplications":"Post-call analysis requires access to recorded control-room audio, transcription, classification and a searchable store; redaction tools work on video, audio and digital evidence within evidence-management workflows.","governanceImplications":"The redaction pilot retained user review and quality assurance; the release does not describe governance for PCA vulnerability flags or how they affect call handling.","securityPrivacyImplications":"Both tools process sensitive call audio and evidence; the release does not describe hosting or data protection arrangements.","caveats":"Awards release, not the evaluation reports; methods, sample sizes and baselines for the PCA and redaction figures are not given; accuracy ratings for redaction are user-reported.","streamIds":["public-safety"],"roles":{"sales":"Sales framing: UK forces are reporting concrete results for AI on call recordings and evidence redaction, with people still reviewing output. That makes measured, human-reviewed productivity tools an easier conversation than autonomous ones. A bounded engagement is a pilot evaluation designed to produce comparable figures locally. This release cannot support claims that the same accuracy or time savings will transfer, since methods and samples are not published here.","engineering":"Engineering framing: both tools depend on clean access to recorded audio and evidence files. Pre-sales should confirm where call recordings and body-camera or interview media are stored, what formats and retention apply, and whether outputs can be written back to records systems. Proof should test transcription and classification accuracy on local calls with a human-labelled sample, and time redaction tasks before and after with reviewer checks.","delivery":"Delivery framing: the work is running a measured pilot with a baseline. Dependencies are access to recordings, a labelled sample, and reviewers to check outputs. Owners are the control-room manager for call analysis and the digital evidence or disclosure unit for redaction. Acceptance should be set in advance against a manual baseline and include error review, not only speed."},"roleDetails":{"sales":{"customerProblem":"Control rooms and evidence units cannot manually review or redact the growing volume of recorded audio and video.","stakeholders":["Control room managers","Digital evidence or disclosure units","Professional standards","Legal services","Chief information officers"],"discoveryQuestions":["What share of calls is reviewed today?","How long does a typical redaction take?","Who checks AI output before release?","Is there a baseline to compare against?"],"valueHypothesis":"Human-reviewed AI analysis and redaction can raise review coverage and cut handling time, if local tests confirm accuracy.","potentialEngagement":"Eight-week measured pilot of call analysis or redaction against a manual baseline.","unsupportedClaims":["That 97.6% accuracy will transfer locally","That redaction accuracy is independently validated","Specific cost savings"]},"engineering":{"fit":"Agencies with recorded call audio and large video or audio evidence volumes.","architectureIntegration":"Connect call-recording and evidence-management stores to transcription, classification and redaction services with write-back.","prerequisites":["Access to recordings","Human-labelled test sample","Records system integration path"],"constraints":["Recording formats and retention vary","UK results may not transfer to U.S. systems","Disclosure rules differ"],"security":"Keep sensitive audio and evidence in approved hosting with access logging.","proofOfValue":["Proposed: compare AI classification with a labelled call sample","Proposed: time redaction before and after with reviewer error checks"]},"delivery":{"work":"Baseline current effort, run the pilot with reviewer checks, and report accuracy and time results.","dependencies":["Recording access","Reviewer availability","Vendor support"],"ownership":"Control room manager or digital evidence unit lead","skillsAdoption":"Train reviewers on verifying AI flags and redactions.","governanceCheckpoints":["Accuracy threshold agreed before pilot","Error review before wider rollout"],"acceptanceCriteria":["Proposed: classification accuracy at or above manual baseline on labelled sample","Proposed: redaction time reduced with no missed redactions in audit"],"risks":["Missed vulnerability flags","Over-reliance on automated redaction","Unpublished evaluation methods"]}},"retrievedAt":"2026-10-22T02:51:53Z","enrichedAt":"2026-10-22T02:56:39Z","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Redaction users reported saving two to four hours per job; Warwickshire paired Copilot rollout with training and peer sessions.","procurementImplications":"PCA was developed with Deloitte; Cleveland piloted a commercial product (Pimloc Redact).","operatingModelImplications":"PCA shifts call review from sampled manual checks to automated analysis of far more calls.","sourceVerification":{"openedUrl":"https://news.npcc.police.uk/releases/national-awards-shine-spotlight-on-the-innovation-transforming-policing","referenceExcerpt":"Evaluation found PCA achieved 97.6% accuracy in identifying key call characteristics, compared with 82% for human reviewers","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}