Operational data with a health score and a debt queue.
Tables is where the estate's working data lives: bases with owners and sensitivity labels, records that show their fill and their agent runs, and a schema-debt queue that ranks the cleanup instead of hiding it.
The overview's first alert is not storage or speed. It is 7 broken link fields, with the fix named: retarget or delete. Data tooling earns trust by showing its own mess first.
tables.cedxsystems.com — live build
Runs on demo data — Northline ops data is the demo workspace, not a customer.
320 records in the sample214 active · 17 locked
24 bases, 625 collaborators48 external shares
502K cell edits in 30 daystable health averages 70
What it is
Three things a data table owes its operators.
Bases with owners, sensitivity and health
24 bases, each with category, table count, records, health, people, external shares, sensitivity and an owner. The incident log holds 892 records at health 93, Restricted; the API keys registry shows 38 people and 5 external shares. Sensitivity is a column, not a promise.
Sensitivity labels: Restricted, Internal, Confidential, Public
320 records with table, base, status, agent run, fill, edits and owner. Lead · Acme 505 sits at 55% fill with an agent run Done; Asset · Sync 878 shows its agent run Failed, with a re-run count of 6 across the workspace. When automation writes, the record says so.
Agent runs on 48 records, 6 needing re-run
Fill percentage and edit count per record
Sensitive-only filter on the table
tables — screen-3
Views are shared surfaces, and counted
120 views across grid, kanban, calendar, gallery and form, each with filters, opens and an external flag. Shared kanban 40 opened 1.2K times in 30 days; 24 views are shared outside the workspace — and the permission blast knows which ones.
24 external views, by link or domain
38K opens in 30 days across the inventory
External-only filter for audit day
tables — screen-4
Product tour
Four screens, captured from the running build.
Not a mockup and not a concept deck. This is what opens at /app/tables.
tables.cedxsystems.com
01 — Overview
Data health, not data vanity
320 records, 96 tables, 103 schema-debt items, 72 share edges and 502K cell edits in 30 days — rolled into a data health of 70 against a 72-to-85 target. The do-this-next queue is scored: revoke the Revenue pipeline's partner share at blast 83, fix the top 7 broken links, archive 51 unused fields.
Records by status: 214 active, 51 draft, 38 archived, 17 locked
41 external share edges, with the blast ranked
Target health printed: 72% → 85%
02 — Bases
The workspace, base by base
From the incident log at health 93 to the vendor register at 58, every base shows what it holds, who is in it and how exposed it is. The ops chip cuts the list to the bases the field teams actually run on.
126 tables claimed across the 24 bases
People and external columns side by side
CSV export of the filtered list
03 — Records
Where the fill rate lives
Records carry a fill bar, an edit count and an agent-run state, so a half-empty record is visible before it poisons a report. Locked records keep their state on screen — 17 of them in the sample.
Search across records, tables, bases, owners
Agent-runs chip isolates automation-written records
Updated dates on every row
04 — Views
Every shared surface, accounted for
Views are the way data leaves the building: 90 of 120 are shared, 24 externally. Each carries its type, filters, opens and owner — the intake form opened 760 times in 30 days and is external. That is the list an audit starts from.
Grid, kanban, calendar, gallery and form types
Opens per view, 30-day window
External flag on every shared row
Who runs it
Three roles keep working data trustworthy.
Roles, not references. We have no named customers yet, so nobody in these photographs is quoted, credited or claimed as one.
Base owners
Run the bases their teams live in — incident logs, training rosters, field-service work orders — and own the health score that comes with the job.
24 bases · 625 collaborators
Data stewardship
Works the schema-debt queue: temp col 3 at 76, the deprecated flag at 72, and 51 archive candidates with null rates above 50%.
103 debt items · 15 high
Automation owners
Own the agent runs writing into records — 48 of them in the sample — and re-run the 6 that failed before the fill rate becomes a lie.
48 agent runs · 6 failed
The shape of it
What the demo workspace actually looks like.
Every figure below is legible in the captures above. Nothing here is a projection of your estate — it is the state of the demo data.
320records in the sample214 active · 17 locked
24bases in the workspace625 collaborators · 48 external
103schema-debt items15 high · queue ranked
502Kcell edits in 30 dayshealth averaging 70
Records by status320-record sample
Active214
Draft51
Archived38
Locked17
Data healthtarget 72% → 85%
70
Average health · 70 across 96 tables
Schema debt 103 · share edges 72 · high debt 15
Views that are shared90 of 120 views
Shared inside the workspace · 90 views
External · 24 views by link or domain
How it runs
Working data's life, in the order it actually happens.
01
Model
Data starts in a base with a category, a sensitivity label and an owner — Restricted, Internal, Confidential or Public, on the row.
02
Fill
Records carry fill percentages and edit counts, and agent runs write too — 48 records in the sample show automation's handwriting.
03
Share
Views multiply the surface: 90 of 120 are shared, 24 of them outside the workspace, and the permission blast scores each edge.
04
Clean
The debt queue ranks the sweep: 7 broken links, 15 high-debt items, 51 unused fields over 50% null — each with the fix named.
One record
The tables the estate already works from.
Tables is not a spreadsheet with logins. It is the structured record the estate's apps, automations and connections read and write.
Finding this out on the third call is worse for you than reading it here, and worse for us.
Tables is not generally available. What opens today is the live build running on demo data — Northline ops data is the software's demo workspace, not a reference.
We have no named customers to show you, so this page shows none. The record names in the captures are the demo data's, not references.
Sensitivity labels — Restricted, Internal, Confidential, Public — are the product's access model, not a compliance certification of any kind.
Record counts, health scores and edit volumes are the demo workspace's telemetry, not benchmarks for yours.
No audit or compliance certification has been issued for Tables. What we can evidence about hosting, encryption and access is on the security page.
Yes. Every screenshot is a capture of the running build and you can open the same build at /app/tables. It runs on demo data from the Northline ops data demo workspace.
What is schema debt?
A scored queue of the cleanup a data estate accumulates: broken link fields, deprecated flags, unused fields over 50% null. The demo workspace carries 103 items, 15 of them high — each with the fix named, from retarget to archive.
Can automation write into records?
Yes, and the record says so. 48 records in the sample show agent runs, with Done and Failed states and a re-run count of 6 — automation's edits are visible, not laundered into the edit count.
How do external shares work?
Views can be shared by link or domain, and 24 are. Each carries an external flag, an opens count and an owner, and the permission blast ranks the widest edges first — the Revenue pipeline partner share scores 83.
Is Tables audited or certified?
No certification has been issued. What we can evidence about hosting, encryption, tenant isolation and production access is written up on the security page.
The workspace is running. Go and look at it.
Live build, demo data, no card. Then count the broken link fields in the base your team actually runs on.