Skip to the page

Solutions

See how it moved. Your systems show today. Keep the days in between.

Many systems only know how things are right now. nexibi keeps a dated copy of a list from a system with a web API, on the schedule you pick, so a board can draw a sprint burndown, or open tickets by status, week by week.

Nothing starts until
you approve it.

Sprint 14Tarnwick Supply Co. · sample trackerSample data
Keep a daily copy of our open issues and chart them by status.
Assistant
Keep a daily copy of Issues
  • Every night at 02:00
  • your time zone
  • keep 90 days
  • counts: by status, by priority
  • burndown rows: per sprint
ApproveChangeNot now
1Only changes are kept
Story points lefttoday · 30 left
0306090Day 02468Day 10 2Each point is a copy kept that night
Open tickets by status
To doIn progressIn review
0204060Wk 38Wk 39Wk 40Wk 4155Wk 42Wk 43Wk 4436Wk 45
  1. 1Only changes are kept
  2. 2Each point is a copy kept that night

Sounds familiar?

In the words product and operations teams use:

  • Our tracker shows today. Ask how the sprint looked last Tuesday, and nobody knows.
  • Someone takes a screenshot of the board every Friday.
  • The weekly status report is rebuilt by hand.

How it works at Tarnwick.

On Tarnwick Supply Co.’s sample tracker.

  1. 1

    Ask, or use the window. Say “keep a daily copy of our open issues and chart them by status”. Or open the four-step window from the connection, or from Query Studio.

    AI Assistant
  2. 2

    Approve one card. The plan is shown on a card: the list, the schedule, the time and time zone, how long to keep, the counts. You can change any of it. Nothing starts until you approve it.

    History engine
  3. 3

    It measures first. How many records there are now, the page size the system really serves, and how many requests each copy costs. Every setting is tried against your system before anything is scheduled, so a wrong project number is refused now, not failed every night.

    History engine
  4. 4

    It keeps only what changed. An unchanged record costs nothing. Each new version records what changed, field by field (“status: In review → Done”). A record that disappears is marked with the time it went, not erased.

    History engine
  5. 5

    Draw it. Ready-made queries come with each copy: every version, as it stands, a daily rollup, a daily breakdown. On a board, stacked columns show one column per copy. Ask the assistant for burndown, burn-up and velocity rows, by sprint and by day.

    Dashboards
  6. 6

    Know when a chart has stopped. The History page says, in plain words, whether each copy is healthy, waiting, failing or paused, and what to do if it is not feeding its charts.

    History engine

3It measures first

Keep a copy of IssuesStep 2 of 4Sample data
Measure
Records now1,284
The system serves100 a page
Requests a copy13
BackNext
Measured before it runs

4It keeps only what changed

Query StudioIssues · every versionSample data
TRK-218 · Export the price list3 versions
Kept_changes
3 Novstatus: Open → In progress
6 Novstatus: In progress → In review
8 Novstatus: In review → Done
What changed, field by field

6Know when a chart has stopped

HistoryTarnwick Supply Co.Sample data
Issuesevery night at 02:00Healthy
Sprintsevery night at 02:30Waiting
Releasesevery night at 03:00Check this

Linked lists, kept as one plan.

  • Keep every project and each project’s issues together: for each project, its issues are fetched and kept linked, up to 200 projects a run.
  • If one project cannot be read, that day gets no bar, rather than a bar missing a project.
  • The History page groups linked lists under a “Plan” badge.

A copy you can trust.

  • It keeps a copy only when you ask for one. Ask a plain question and you get the answer, with a one-line offer of a copy.
  • A run that cannot read the whole list stores nothing that day, rather than a short bar.
  • A field that vanished from the system is reported, not charted as “(none)”.
  • When every record changes at once, it warns you: that is more likely a settings change than real change.
  • Pressing “Check now” never doubles a bar.
  • After a system has been down, it runs one catch-up, never a burst.
  • A scheduled copy never changes data in your system.

What you get.

  • A dated copy of the lists you choose, on your schedule and in your time zone.

  • Boards that show how things moved: a sprint burndown, tickets by status, week by week.

  • What changed, field by field, and when.

  • When a copy brings new data, the forecasts and predictions that read it update by themselves.

History engine

Before you start.

Twelve weekly columns
take twelve weeks.

  • It keeps lists from systems with a web API, the ready-made connectors included. Databases, files and mailboxes cannot be kept.
  • The history starts on the day you turn the copy on. Twelve weekly columns take twelve weeks.
  • A copy reads up to 50,000 records a run.
  • Detailed versions are kept 30, 90 or 365 days. The daily counts and burndown rows are kept beyond that, so a 90-day setting still keeps a two-year chart.
  • The copy is kept in your company’s own history database, and seen by whoever its connection is shared with.
  • A system that needs someone at a browser to sign in cannot be copied with nobody watching. Where each person signs in as themselves, the copy uses its owner’s own sign-in.

What it does not do yet.

So you know before the demo.

  • It cannot fill in the past.
  • A board card cannot yet show a list as it stood on a past date, beside today.
  • Asking the chat how a figure changed over time is not ready yet.Draw it on a board.

For product teams, see Product. Product

Bring the chart your tracker cannot draw.

In the demo we connect one of your systems, and you ask the questions.