Deflect support tickets only when a calibrated confidence score is high, and queue the middle band for human review
An open-source ticket deflection engine escalates risky tickets directly, answers procedural ones by rule, and uses retrieval over documentation with three confidence bands for auto-reply, human review and escalation.
Evidence: The author reports this. We have not checked it beyond reading the source.
The business problem
Automating support replies saves time, but a confident wrong answer on a billing or security issue is costly.
What was tried
A security and SLA check sends high-severity tickets (billing disputes, account compromise, breach reports) straight to a second-line engineer. An exact-match handler answers common procedural requests such as password resets and invoice downloads. Other tickets go to retrieval over verified documentation, where a re-ranker produces a calibrated probability: 0.85 or higher is auto-replied, 0.65 to 0.85 is drafted for a human to approve, and below 0.65 goes to the queue with a timer.
What was reported (positive)
The README claims 70% deflection and 90% decision accuracy on verified ticket sets, and 0.0% false deflection on security and high-risk billing intents.
Limitations
Results are self-reported on unspecified data, the corpus size, accuracy definition and reproducibility are not stated, and the 0% figure covers only two intent types. The models and ticketing integrations are not named, there is no limitations section, and the repository is very new (nine commits, no stars or releases). It is MIT-licensed.
What you need
Python 3.11 or later and a verified documentation set. The embedding, re-ranking and language models are not named.
Sources
- GitHub (nathaniel-gordon/deflectiq) ↗ Code repository, publication date unknown
Source published: unknown. Last reviewed here: October 11, 2026. Spot a mistake? Tell us.
Related workflows
Roll out OpenClaw for inbox triage, meeting notes and CRM updates in phases with approvals and isolation
An infrastructure vendor's guide sets out business uses for OpenClaw and a phased rollout: pilot a low-risk workflow, add guardrails and approvals, then scale once value and safety are shown.
Measure true resolution, not just deflection, for an AI support agent with an escalation policy
A small open-source evaluation shows a naive support agent deflecting every ticket while resolving about half, and a policy-gated agent that deflects fewer but resolves what it handles.
Draft help-centre articles automatically from resolved tickets and keep a human in charge of publishing
A merged helpdesk change that, when a ticket is resolved, asks an AI whether other customers are likely to ask the same thing and writes a draft article for managers to review if none exists.
Draft help-centre-cited support replies and propose refunds that a person must approve
An open-source support workbench classifies tickets into three tiers, drafts replies that cite help-centre articles, reads billing data in read-only test mode, and only proposes refunds that need a named approver.