Overview

Tickets, incidents and partner issues from a service desk, an incident platform and a partner channel, synthetic by default, plus a handful of real incidents captured once from the owner's own Rootly and Jira accounts, unified into one queue. A personal portfolio demo, not a production tool.

  • Correlated automatically. The same underlying problem reported in two different tools gets linked here with a confidence score and the evidence behind it, not just asserted. Real records from Jira and Rootly were linked this way at 97% confidence. See it linked →
  • Short on time? A two-and-a-half-minute narrated walkthrough covers every screen. Watch it →
  • SLA scored per audience. Internal, consumer and partner get separate scorecards and separately-tiered targets, never one blended number. SLA health →
  • Mostly synthetic, a little real. Rows marked Live are genuinely captured from the owner's own trial accounts, replayed rather than fetched live, and never correlated with the invented signals.
  • Everything here is computed, not scripted. Move the threshold slider, edit an SLA target, or add your own source, and the numbers recompute live in this browser. Connect a source →
  • Nothing you change leaves this browser. Dismissals, SLA edits and the change history are local only, hash-chained so what's logged can be checked, not just trusted.
  • Where to go next. This Queue is every signal in one place; Correlation explains why some got linked and some didn't; SLA health scores the three audiences separately; Connect a source proves the architecture is real. For the full build story, the tech stack and every test that passed, see About →

Unified queue

One list across every connected tool. Each signal keeps its source badge; where separate tools are describing the same underlying problem, the row links to the correlation that explains why.

Signals

31

from 6 sources

Correlation groups

5

likely shared root causes

Signals in a group

15

of 31

Standalone

16

deliberately left unlinked

13 real records captured from Jira and Rootly (latest capture 2 Oct 2026, 11:11 UTC) sit alongside the synthetic signals, marked Live. Real and synthetic signals are never correlated with each other.

Showing 4 of 31

  • Partner ChannelPC-302

    Halcyon Travel: booking applications timing out at final step

    Halcyon Travel · partner portal · open 2h 30m

    Linked to change CHG-1006 (normal, release the redesigned application form), deployed 2h 40m before · 60% confidence

    Supplier (invented): Hollowfen Hosting

    PartnerMediumOpenLinked · 5 signals
  • Incident PlatformINC-202

    Synthetic check failing: partner portal submit endpoint p95 above 8s

    partner portal · open 2h 40m

    Linked to change CHG-1006 (normal, release the redesigned application form), deployed 2h 30m before · 63% confidence

    Supplier (invented): Hollowfen Hosting

    InternalMediumInvestigatingLinked · 5 signals
  • Partner ChannelPC-301

    Meridian Retail: customer applications taking minutes to submit

    Meridian Retail · partner portal · open 2h 52m

    Linked to change CHG-1006 (normal, release the redesigned application form), deployed 2h 18m before · 66% confidence

    Supplier (invented): Hollowfen Hosting

    PartnerHighOpenLinked · 5 signals
  • Service DeskSD-101

    Users reporting slow application submission on partner portal

    partner portal · open 3h 20m

    Linked to change CHG-1006 (normal, release the redesigned application form), deployed 1h 50m before · 73% confidence

    Supplier (invented): Hollowfen Hosting

    InternalHighInvestigatingLinked · 5 signals