Power Tables App Overview

Apps Getting Started Updated July 20, 2026 17 views

Power Tables App Overview

Store and query structured data at database scale — typed tables that hold millions of rows, with fast search, saved views, CSV and Excel import and export, scheduled imports, validation rules, row history, a REST API, and a direct feed into Live Boards, dashboards, and Digital Signage.

What is the Power Tables App?

Power Tables is the MangoApps data app for structured records at scale. Administrators create typed tables — every column has a real data type such as number, currency, date, or select — import CSVs, and query the data at database speed. Each Power Table is backed by a real database table, so scale is bounded by indexes, not by what fits in a browser: a table with a few million rows sorts and filters as quickly as one with a few hundred.

Power Tables is deliberately not a spreadsheet. There are no cells to click into, no formulas, and no free-form grid editing. The grid is a fast read-only viewport; adding or changing data happens through a typed row form — the way a database UI edits records, not the way a spreadsheet edits cells.

Core Value Proposition:

Database scale — One real, typed, indexed table per Power Table; millions of rows without slowdown
Typed columns — Eleven column types validate data on the way in, so a date column only ever holds dates; per-column validation rules add required, min/max, and format-pattern checks on top
Fast querying — Full-text search, per-column filters, and index-backed sorting at any depth
Saved views — Capture a filter/search/sort combination once and reuse it everywhere, including as the data source for Live Boards
Bulk import and export — CSV and Excel imports up to 100 MB, with an explicit column-mapping step, run as tracked background jobs with live progress; any view exports back to CSV or Excel with a download when it’s ready
Automated and programmatic access — Per-table scheduled imports pull a CSV or Excel file nightly from a public https URL, an authenticated endpoint, or an SFTP server, and a REST API reads and writes rows with least-privilege token scopes

The data trio: AI Sheets (work with data — spreadsheet editing, formulas, up to 50,000 rows) · Power Tables (store data at scale — typed records, millions of rows) · Live Boards (show data live — leaderboards and KPI tiles on dashboards and signage).

At a Glance

Setup Time	Column Types	Row Capacity	Feeds
Under 2 minutes	11 typed	Millions per table (quota-configurable)	Live Boards, dashboards, Digital Signage

Perfect For:

Operations teams — Store performance data, asset registers, inspection results, and production metrics at full history depth
Sales and frontline leaders — Keep the numbers that drive leaderboards (sales by store, units by rep) in one governed place
Administrators — Replace the “giant spreadsheet nobody can open anymore” with a table that stays fast forever

Key Concepts

Concept	What it is
Table	A named collection of typed columns and rows, backed by its own real database table. Tables are created, archived, and restored by admins.
Column	A field with a fixed type (text, long text, number, currency, boolean, date, date & time, select with options, email, URL, People). A column’s type is fixed after creation — add a new column to change types.
Row	One record. Rows are added and edited in a typed drawer form — date pickers for dates, dropdowns for selects, Yes/No for booleans.
View	A saved combination of filters, search, and sort over a table. Views are the binding target for Live Boards — boards read views, never raw tables.
Index	A per-column speed structure built in the background. Indexed columns sort and filter instantly at any table size.

Power Tables vs AI Sheets

Both apps hold tabular data. They solve different problems, and knowing which to reach for saves rework:

	AI Sheets	Power Tables
Mental model	Spreadsheet — edit cells, collaborate live	Database — edit records, query at scale
Row capacity	Up to 50,000	Millions (default quota 1,000,000 per table, configurable to 10,000,000)
Editing	Click any cell, type, autosave	Click a row to open a typed edit form
Formulas	Yes — per-row formula columns	No — by design
AI features	Column generation, natural-language queries, AI edits	Optional Ask AI agent (admin toggle, off by default) for questions, filtered aggregates (counts, sums, averages), and confirm-gated actions
Who edits	Any collaborator (Owner/Editor)	Admins and Power Tables app admins
Best for	Team trackers, checklists, ad-hoc analysis	Reference data, metrics history, leaderboard sources, anything that outgrows a sheet

If your data needs formulas, cell-level collaboration, or AI analysis, use AI Sheets. If your data needs to be big, typed, and fast — and to feed Live Boards — use Power Tables.

Key Features

Typed Tables

Create a table from a drawer form with a name, description, and up to four initial columns (more can be added any time). The Start from a template picker offers four prebuilt schemas — Asset registry, Visitor log, Store performance, and Incident tracker — that create the table with its typed columns already in place (the manual column rows are ignored when a template is chosen), so a new table starts on structure instead of blank. Every column gets one of eleven types:

Type	Behavior
Text / Long text	Free text; long text gets a larger input
Number	Numeric values, validated on write
Currency	Numeric with currency display formatting
Boolean	Yes/No
Date / Date & time	Real date pickers in the row form
Select	A fixed option list you define (comma-separated when creating the column)
Email / URL	Text with format validation
User (People)	References a person in your business — picked with the people search in the row form; names display in the grid and in CSV exports

Because columns are typed, bad data is rejected at the door — a number column never silently accumulates “N/A” strings.

Validation Rules per Column

Beyond the type itself, each column’s form has a Validation rules section:

Required — rows must have a value in this column
Minimum / Maximum — numeric bounds for number and currency columns
Format pattern — a regular expression that text values must match (for example \A[A-Z]{2}-\d{4}\z for an asset-tag format)

Rules are enforced on every row write — the row drawer, bulk edit, the REST API, and per-row on imports, where a failing row is reported with its row number and message while the rest of the file still lands.

The Grid: Read Fast, Edit Precisely

The table page shows a high-performance grid that streams rows from the server as you scroll — the browser only ever holds the visible window, which is why millions of rows stay smooth.

Full-text search across the table’s text columns from the search box
Per-column filters — text, number, and date columns each offer type-appropriate filter operators
Sorting on any column; sorting an unindexed column on a large table (10,000+ rows) shows a nudge suggesting you index that column
Row counts are estimates on large tables (shown as “~1.2M rows”) so the page never blocks counting

Editing is row-level: click any row (admins only) to open the row edit drawer with one typed input per column — including a people search for User columns — plus a Delete Row button. Add Row opens the same drawer blank. If someone else changed the row while you were editing, the save is stopped, your typed values are kept, and the form tells you to review and “Save again to overwrite” — no silent lost updates. Each row also has a pinned trash icon for quick deletion, and admins can tick the row checkboxes and use Delete Selected to remove many rows at once as a tracked background job.

Row History

The row edit drawer includes a collapsible History panel showing the row’s audit trail: when it was added, when it was deleted, and every field change as old value → new value with who made it. Entries are kept for 180 days. Imports and bulk inserts are deliberately not logged row-by-row — for imported data, the import task record is the audit trail.

Bulk Edit

Select rows in the grid and a bulk-edit bar appears alongside Delete Selected: pick a column and a value, then click Set on N. When the chosen column is a People column, the free-text value box swaps for the canonical people picker, so you assign a person the same way the row drawer does. Leaving the value blank (or picking no one) clears the column. After a confirmation, the update runs as a tracked background job — each row goes through the same typed casting and validation rules as the drawer, so bad values fail per row with a message on the task card while the rest of the selection updates. Selections are processed up to 1,000 rows per batch, and anything beyond that is disclosed, never silently dropped.

Saved Views

Admins can capture the grid’s current filters, search, and sort as a named view from the Views dropdown. Views appear in the dropdown for every user, alongside the built-in “All rows”. The active view can be updated with the current filters or deleted — with one guardrail: a view that a Live Board is bound to cannot be deleted until the board is repointed or archived.

CSV and Excel Import

The Import CSV or Excel drawer accepts .csv and .xlsx files up to 100 MB (enforced in the browser and again on the server; for an Excel workbook the first sheet imports). Importing is a two-step flow: after the upload, a Map Columns screen lists every file header so you decide where each one lands — an existing column (pre-selected when the names match), a new text column, or skipped entirely. Nothing imports until you start it. Imports then run as background jobs with a live progress card shown above the grid; failed imports stay visible with their error details rather than disappearing. Imports stop at the table’s row quota and report how many rows were skipped.

Scheduled Imports

Each table can have one scheduled import — configured in the “Scheduled import” card on the table page — that pulls a CSV or Excel file every night from a public https URL, an authenticated endpoint, or an SFTP server:

Replace rows — the table’s rows are cleared and re-imported each run; use this when the source system is the system of record
Append rows — every run adds rows; use this for accumulating log-style sources
Run now — trigger the pull on demand without waiting for the nightly run
The card shows the last run’s time and status; a failed run shows its error right there and the card opens expanded

The nightly run uses the same import pipeline as a manual upload — name-based column mapping with new headers created as text columns, per-row error reporting, and a progress card.

Sources behind a login work too. An admin configures the connection once at Admin > Integrations (/admin/integrations), and the schedule card’s Connection select picks it up: a Data Feed connection sends its credential — a bearer token, basic auth, or a custom header — with the nightly HTTPS pull, and an SFTP connection swaps the URL field for a remote file path on the SFTP server. Plain public https URLs keep working with no connection at all (“None — public URL” is the default).

CSV and Excel Export

The Export CSV and Export Excel buttons capture the grid’s current state — filters, search, sort, and the active view’s visible columns — and run the export as a background task. When it completes, the task card above the grid offers a Download button; People columns export as names. Export what you see, not a raw dump. Excel export is capped at 100,000 rows (use CSV, which has no cap, for bigger tables — the export refuses cleanly rather than truncating silently).

Columns Panel and Indexes

An expandable Columns panel on the table page (admins only) lists every column with its type and index state:

Add / Edit — add columns any time; edit renames a column or updates a select column’s options
Index / Unindex — index builds run as tracked background tasks (“Building…” then “Indexed”); failures surface on the task card instead of vanishing
Archive — retires a column from the table

Table Lifecycle: Archive with a Grace Period

Archiving a table hides it (and any boards reading it) immediately, but the data survives a 7-day grace period during which the table can be restored intact. After the grace period the underlying data is permanently purged; the table then shows as “Purged” and can no longer be restored.

Quotas and Settings

Admins tune four quotas in Settings: max tables (default 10), max rows per table (default 1,000,000), max columns per table (default 50), and max indexed columns per table (default 8). Leaving a field blank keeps the stored value. When the table quota is reached, the New Table button is disabled with an explanation rather than failing after you fill in the form.

Settings also holds the “Enable the Power Tables AI agent in Ask AI” toggle (off by default). When enabled, Ask AI can answer questions about your tables — row counts, table details, row lookups, import activity, and filtered aggregates: ask “total revenue for the North region” or “average handle time this week” and the agent computes the count, sum, average, minimum, or maximum directly against the table, with currency results formatted as money. Writes — create a table, save a view, start a column index — always go through an explicit confirmation step.

REST API

The public REST API at /api/v1/power-tables makes tables available to scripts and integrations:

Endpoint	What it does
`GET /api/v1/power-tables`	List active tables with row estimates
`GET /api/v1/power-tables/:id`	A table’s structure — columns (with types, options, required flags) and views
`GET /api/v1/power-tables/:id/rows`	Row reads with the grid’s filter grammar, search, sort, and keyset pagination (opaque cursor, up to 500 rows per page)
`GET /api/v1/power-tables/:id/aggregate`	Filtered count, sum, avg, min, or max over a column
`POST /api/v1/power-tables/:id/rows`	Ingest one row — same casting, quotas, validation rules, and audit trail as the row drawer

Access is by API token with least-privilege scopes: read:power_tables for the read endpoints, write:power_tables for ingest — and writes additionally require the token’s user to be a Power Tables admin, exactly like the web surface.

Settings also offers Load Sample Data — a Store Performance table with 48 rows, a saved view, and two Live Boards, so you can see the whole pipeline working and remove it with one click.

User Roles & Permissions

Role	Capabilities
Admin (business admin or Power Tables app admin)	Everything: create/edit/archive/restore tables, add/edit/index/archive columns, add/edit/delete rows, import CSV, manage views, configure settings, load sample data
Member with app access	View tables, open the grid, search, filter, sort, and use saved views — read-only
User without app access	No access; the app does not appear

Left sidebar: Apps > Power Tables (direct URL: /apps/power-tables)
The tables list shows Active / Archived / All filter chips, a quota subtitle (“N/M tables used”), and pagination
Click a table name to open its grid; Settings is in the page header (admins)
On a phone, /m/apps/power-tables offers a read-only mobile view — table cards and searchable row cards; importing, exporting, and editing stay on desktop

Getting Started

The end-to-end workflow (table to live display)

Create a table — Apps > Power Tables > New Table. Name it (for example “Store Performance”) and add initial columns with the right types (Store as text, Units Sold as number, Date as date).
Import your data — open the table, click Import CSV, choose your CSV or Excel file, map each file column on the Map Columns screen, and watch the progress card above the grid.
Shape a view — filter and sort the grid to the slice you care about (for example this quarter, sorted by units sold), then open Views and use Save view to name it.
Feed a board — in the Live Boards app, create a board bound to that saved view. The board appears on dashboards and, if published, on Digital Signage — and it stays current as the table’s data changes.

For Administrators

Enable Power Tables from the Apps Marketplace (it is a pay-as-you-go app — you opt in; billing follows actual rows stored)
Review Settings quotas for your expected data volume
Use Load Sample Data to explore the pipeline before importing real data
Index the columns you sort and filter on most

Pricing

Power Tables is a pay-as-you-go app billed on actual usage, not per seat: $1.20 per 1,000 rows per year for the first 1 million rows, $0.36 per 1,000 rows from 1M to 10M, and $0.24 per 1,000 rows beyond that. Rows are counted nightly across active tables; archived tables do not bill. There is no floor and no minimum — importing a 10,000-row spreadsheet costs about $12 per year. The app is off until an admin enables it.

Best Practices

Type columns correctly up front — a column’s type is fixed after creation; add a new column if you need a different type
Start from a template when one fits — the four prebuilt schemas land you on a correctly typed table in one click
Add validation rules to columns that gate data quality — required, min/max, and format patterns catch bad values at write time, including per-row on imports
Index your sort/filter columns — indexes are what keep big tables instant; the grid nudges you when a sort would benefit
Save views for every recurring slice — views drive Live Boards and keep filter logic in one governed place
Archive instead of deleting — you get a 7-day safety window, and archived tables stop billing
Review the Map Columns screen on every import — unmatched headers default to “Create as new text column”; pre-create typed columns and map to them so values validate on import, and switch stray headers to Skip

Frequently Asked Questions

Q: How is this different from AI Sheets?
A: AI Sheets is a collaborative spreadsheet — cell editing, formulas, AI analysis, up to 50,000 rows. Power Tables is a database — typed records, row-form editing, millions of rows, and the data source for Live Boards. See the comparison table above.

Q: Who can add or edit data?
A: Business admins and Power Tables app admins. Members with app access can view, search, filter, and use saved views, but the grid is read-only for them.

Q: What happens when I archive a table?
A: It disappears from the active list (and any Live Boards reading it show “Source unavailable”) immediately, but the data is kept for 7 days, during which you can restore it. After that the data is permanently purged.

Q: Can I export my data, use formulas, or access the data via a public API?
A: Export, yes: Export CSV and Export Excel capture the current view — filters, search, sort, and visible columns — and deliver the file through a background task’s Download button (Excel is capped at 100,000 rows; CSV has no cap). API, yes: the REST API at /api/v1/power-tables reads tables, rows, and aggregates and ingests rows under read:power_tables / write:power_tables token scopes. Formulas and per-table permissions are not in this version — deliberate scope decisions; for formula workflows, use AI Sheets.

Q: Can imports run automatically?
A: Yes — each table can have one scheduled import that pulls a CSV or Excel file nightly, in replace or append mode, with a Run now button for on-demand pulls. The source can be a public https URL with no setup, or — through a connection configured once at Admin > Integrations — an authenticated endpoint (bearer token, basic auth, or custom header via a Data Feed connection) or a remote file path on an SFTP server.

Q: Is there a record of who changed a row?
A: Yes, for single-row changes: the row edit drawer’s History panel lists every field change (old value to new value, with the actor) plus row adds and deletes, kept for 180 days. Imports and bulk inserts are not logged row-by-row — the import task record is the audit trail for those.

Q: Why did my two saves conflict?
A: Power Tables uses optimistic concurrency: if another admin saved the same row while your edit drawer was open, your save is held, your typed values are preserved, and you are asked to save again to overwrite the newer version — so nobody’s changes are silently lost.

Q: How big can a table get?
A: The default quota is 1,000,000 rows per table, and admins can raise it to 10,000,000 in Settings. Search, filters, and indexed sorts remain fast at that scale because the work happens in the database, not the browser.

Apps & Extensions Overview — Browse all MangoApps marketplace apps
Live Boards App Overview — Turn Power Tables views into live leaderboards and KPI tiles
AI Sheets App Overview — Collaborative spreadsheets with formulas and AI, for data under 50,000 rows
Digital Signage App Overview — The screens your boards can publish to

Give your data a real backbone — typed, indexed, and ready to power every live display in your workplace.

Was this article helpful?

Article ID: power_tables_app_overview

Power Tables App Overview