{"resourceId":"runshaw-college-bobbie-ai-ticket-triage-september-peak-20261022","versions":[{"version":"external-ac9b4b98e138eec81a38607630cdda2ed707e37e4e4f3b5ceae1880147562aed","resource":{"id":"runshaw-college-bobbie-ai-ticket-triage-september-peak-20261022","title":"Runshaw College's IT head reports AI-assisted ticket triage at about 95% accuracy through a September peak of 3,240 tickets, with every result human-reviewed","organization":"Runshaw College (UK further education), IT Services","sector":"Further education IT service management","geography":"United Kingdom (Lancashire, England)","publishedAt":"22 October 2026 (Service Desk Institute resource)","publicationDate":"2026-10-22","eventDate":null,"sourceName":"Service Desk Institute","sourceLabel":"Practitioner case write-up by the college's Head of IT Services on an industry-body site","sourceUrl":"https://www.servicedeskinstitute.com/resources/ai-that-supports-people-runshaw-college/","evidenceClass":"public-sector-association","outcomeClass":"effective","topics":["developers-agents","operating-model","accessibility-workforce"],"finding":"Alex Harding, Head of IT Services, reports that in September 2026 the service desk handled 3,240 tickets, 23% more than the previous September, with average triage time of 7.89 minutes, 99.88% triaged within target and CSAT of 4.99/5. An IT Services agent, 'Bobbie', built on Microsoft Copilot capability and Jira Automation, suggests category, assignment, title and summary; it was trained on a year of previous tickets and is reported at around 95% accuracy. Every result is human-reviewed and team leaders control the rules. Time released was described as equivalent to about two staff, without removing people. Bobbie also reviews proposed changes before the Change Advisory Board, and AI-assisted problem management clusters incidents every few hours for the team to decide on action.","sledRelevance":"Directly transferable pattern for campus IT service desks facing start-of-term peaks: AI suggests triage fields inside the existing ITSM tool, humans confirm, and gains are framed as released capacity rather than headcount reduction.","evidence":"Self-reported operational metrics from the institution's IT head; no baseline triage time, no accuracy method, no independent audit. Published on the Service Desk Institute site (an industry body that also audits the college's service desk).","architectureImplications":"Agent built from existing Microsoft Copilot, Jira Automation and Microsoft Foundry AI within the current estate; writes suggested fields back into Jira; problem clustering routes summaries to a Teams channel.","governanceImplications":"College AI guardrails applied; team leaders own triage rules; AI does not decide root cause or raise Problem records. Mapped to SDI Global Best Practice Standard v9 Concept 10.","securityPrivacyImplications":"Model trained on a year of historical tickets, which can contain personal data; the article does not describe data minimisation, retention or access controls.","caveats":"Single institution, self-reported and published by an industry body with an audit relationship; '95% accuracy' and 'two staff' equivalence are not methodologically explained; incident-ratio improvement since 2016/17 has several contributors.","streamIds":["campus-operations"],"roles":{"sales":"Sales framing: peak-period service desk capacity is a recurring pain that AI triage can relieve without a headcount story. Runshaw reports about 95% triage accuracy and roughly two staff-equivalents of released time across a 3,240-ticket September, using tools it already licensed. Interpretation: lead with a bounded triage pilot inside the existing ITSM platform, framed around start-of-term peaks and human review, and avoid promising savings the college itself does not quantify in money.","engineering":"Engineering framing: this is suggestion-mode classification written back into Jira, not autonomous resolution. Validate with the customer's own labelled ticket history, measure accuracy per category against human final routing, and confirm the platform can log suggestions versus accepted values. Historical tickets carry personal data, so training and retention need review. Copilot, Jira Automation and Foundry dependencies are specific to this estate; other ITSM platforms need equivalent hooks.","delivery":"Delivery framing: the operating change is that team leaders own the triage rules and every suggestion is checked. Delivery should set up a shadow period, a correction-feedback loop and peak-period reporting on triage time, within-target rate and satisfaction. Train desk staff on reviewing suggestions rather than re-keying. Guard against automation complacency: track override rates, and keep problem-candidate decisions with the team as Runshaw does."},"roleDetails":{"sales":{"customerProblem":"Start-of-term ticket peaks overwhelm triage capacity on small college service desks.","stakeholders":["Head of IT Services","Service desk team leaders","CIO/IT director","Student services leads"],"discoveryQuestions":["What share of desk time goes to categorising and assigning tickets?","Do you have a year of reliably labelled tickets?","Which AI capabilities are already licensed in your ITSM and productivity suite?"],"valueHypothesis":"AI-suggested triage may release desk capacity during peaks; Runshaw reports about two staff-equivalents but no monetary saving.","potentialEngagement":"A one-term shadow-then-assist triage pilot in the existing ITSM tool with human review.","unsupportedClaims":["Guaranteed headcount or cost savings","95% accuracy on another institution's tickets","Autonomous ticket resolution"]},"engineering":{"fit":"Colleges with a mature ITSM tool, consistent categories and Copilot or equivalent automation already licensed.","architectureIntegration":"Classifier suggests category, assignee, title and summary via ITSM automation; results written to ticket fields; problem clusters posted to a collaboration channel.","prerequisites":["Labelled historical tickets","Stable service catalogue categories","ITSM automation and API access"],"constraints":["Estate-specific Microsoft/Atlassian build","No published accuracy method","UK FE context"],"security":"Review personal data in historical tickets used for training and restrict access to model outputs and logs.","proofOfValue":["Per-category accuracy versus human final routing","Override rate during shadow period","Triage time before and after on a comparable peak"]},"delivery":{"work":"Configure suggestion-mode triage, run a shadow period, then enable assisted triage with leader-owned rules and reporting.","dependencies":["ITSM admin access","Clean category taxonomy","Desk staff time for review"],"ownership":"Service desk team leaders own rules; Head of IT Services owns outcomes.","skillsAdoption":"Train agents to review and correct suggestions and feed corrections back.","governanceCheckpoints":["Data protection review of training tickets","Go/no-go after shadow accuracy review","Termly review of override and error rates"],"acceptanceCriteria":["Proposed: at least 90% suggestion acceptance on in-scope categories in shadow mode","Proposed: triage-within-target rate maintained at peak","Proposed: no reduction in satisfaction score"],"risks":["Automation complacency in human review","Drift when categories change","Overstated savings claims"]}},"retrievedAt":"2026-11-07T21:49:30-06:00","enrichedAt":"2026-11-07T22:00:00-06:00","enrichmentBasis":"retrieved source","accessibilityWorkforceImplications":"Explicitly positions AI as releasing time for customer work during peak demand rather than reducing staff.","procurementImplications":"Uses capabilities already within the Microsoft and Atlassian estate rather than a new point product.","operatingModelImplications":"Human-in-the-loop triage with leader-owned rules, scheduled problem-cluster reviews and pre-CAB change review shifts staff time from sorting to resolution.","sourceVerification":{"openedUrl":"https://www.servicedeskinstitute.com/resources/ai-that-supports-people-runshaw-college/","referenceExcerpt":"Every result remains subject to human review, and Service Desk Team Leaders control the rules used by the process.","promptVersion":"sled-research-v4.0","model":null,"basis":"agent-reported inspection"}}}]}