ATX AnalyticsTable Documentation | Every Button, Dialog and Setting

ATX AnalyticsTable documentationThe complete functional reference for ATX AnalyticsTable 1.6.0: 17 sections, 123 pages and about 150 thousand words covering every ribbon tab, button, dialog and setting, with step-by-step click paths such as Transform → Pipeline → Pipelines. Click paths read Tab → Group → Button → control inside the window that opens, exactly as labelled in the English interface. Every page is outlined publicly; the full text is available to signed-in users, and a few sample pages are open to everyone.

Getting started

How this documentation is organised, the ribbon, the three kinds of data a cell can hold, and the project model.
About this documentation — What is covered, who it is for, and how to read a click path.
The workspace and the ribbon — The ribbon tabs and groups from left to right, and how dialogs, panels and menus behave.
Data modes: grid, preview pin and static pin — The three kinds of data a cell can hold, and why a command can behave differently on each.
Projects, saving and the project folder — Create, open, save and close projects; what is stored in the project folder.

Installation & deployment

Requirements, the Windows and Linux packages, server deployment, fleet policy, sign-in and roles, single sign-on, SCIM provisioning, revoking access, corporate networks, and monitoring.
System requirements — Supported platforms, hardware, disk space and what each kind of installation needs.
Installing on Windows — The four Windows packages, silent installation, port, licence, proxy and theme options, upgrades and uninstalling.
Installing on Linux — AppImage, .deb and .rpm packages, the install script, running as a systemd service, and platform support.
Server deployment: Docker, Compose and Helm — The container image, the rules every server deployment must satisfy, the Compose stack, the Helm chart and the vault key.
Fleet configuration and policy — Group Policy and Intune templates, config.ini, machine-wide policy files including the AI model policy, and which source wins.
Sign-in and roles — Entra ID sign-in and its three switches, the viewer, analyst and admin roles, how roles are mapped — from Entra, OIDC, SAML or SCIM — service tokens and project roots.
Single sign-on with OpenID Connect and SAML 2.0 — Sign people in through Okta, Auth0, Keycloak, Ping, OneLogin, Google, ADFS or any other OpenID Connect or SAML 2.0 identity provider — what is verified, how sessions work, and how to set it up.
SCIM provisioning — users and groups from your directory — The SCIM 2.0 endpoint your identity provider pushes users and groups to, the authoritative directory that decides roles on every request, and how to connect Entra ID and Okta.
Revoking access — stop a person now — End a person's sessions and refuse every token issued before now, block a leaver until unblocked, and bound how long a removed role can survive with a maximum token age.
Corporate networks: proxies, TLS inspection and outbound controls — Proxy selection and PAC, proxy sign-in, TLS-inspection certificates, the server's SSRF guard and the database TLS floor.
Monitoring and support — Prometheus metrics, health endpoints, request ids, logs, the diagnostics bundle, crash dumps and graceful shutdown.

Settings tab

The in-app manual, every section of the Settings window, settings export and import, project parameters, accessibility, interface languages and semantic objects.
Settings tab — controls — Every control on the Settings tab: Manuals, Set / Restore, Export, Import, Parameters, Dark / Light and Reduce Motion.
Application settings reference — Every section of the Settings window, in the order the window lists them: what each holds, who sees it, and its fields.
Project parameters — Named values referenced as ${name} in SQL, file names, paths and forecasts — typed, or read from a cell.
Interface languages — Switching language, the six languages that ship with the app, installing a language file without reinstalling, and writing a translation.
Semantic layer — Governed business definitions — datasets, dimensions with hierarchies, measures, calculated members, joins, filters and parameters — shared across the installation or kept in a project, certified by data stewards, queried with Try a query, and used by Ask Greg.

File tab

Project lifecycle, sharing, the scheduler, backups, cloud copies and navigation.
File tab — controls — Every button on the File tab.
Scheduler — unattended refresh, imports, pipelines and exports — What can be scheduled, triggers, execution order, retries, notifications and the run history.
Backup and restore — Versioned project backups to a folder, and restoring a backup into a new folder.
Go To, Find and Replace — Jumping to a cell and searching or replacing across sheets.

Design tab

Cell formatting, copy and paste, conditional formatting, merging and grouping, borders, cell properties, table design and hyperlinks.
Design tab — controls — Every group on the Design tab.
Copy, cut and paste — Copy with or without formatting, paste values, transpose, and what a paste carries.
Conditional formatting — Stoplights, heat maps, colour scales, data bars, icon sets and the rule manager.
Merging cells and grouping rows or columns — Merge, merge across, merge down, unmerge; outline groups and collapse behaviour.
Table tab — designing a database table — Declaring column types and constraints in the grid, value checking and the DDL created per engine.
Hyperlinks — Linking cells to web pages, other sheets and other projects; following links, export, security and the link manager.

Transform tab

Sort and filter, reshape columns, PII masking, pipelines, alerts, value lists and data quality.
Transform tab — controls — Every button on the Transform tab.
Sort and filter — Quick and multi-level sorts, column filters, and how filters are saved and restored.
Reshape: Text to Columns, Merge Columns, Value Map, Dedupe — Splitting and joining columns, recoding values with a rule table and removing duplicate rows.
Pipelines — repeatable data preparation — Create, bind, build, run, step through and schedule a pipeline; every step type and its options.
Alerts — Rules that watch the data and notify when a condition holds.
Value lists (validation) — Dropdown lists for cells: sources, caching, refresh and assignment.
Mask PII — personal data protection — Scanning for personal data and masking it by redaction, partial masking, hashing or synthetic values.
Anomalies — invisible and malformed text — Finding and fixing line breaks, exotic spaces, zero-width characters, markup and export metadata.
Outliers — statistical detectors — Numeric, multi-column, categorical and date detectors, reading results, and writing the report onto a sheet.

SQL tab

The SQL editor, the query library, the visual query builder, Git and query flow.
SQL tab — controls — Every button on the SQL tab.
Visual query builder, query library and monitor — Building queries without SQL, saving them to the library, Git and monitoring running queries.

Get Data tab

Data sources, connections, database, file and cloud imports, and import history.
Get Data tab — controls — Every button on the Get Data tab.
Connections — Creating and managing database, file and cloud connections and their credentials.
Connecting to databases and warehouses — Choosing a sign-in method, and the address, sign-in and limits for Snowflake, Databricks, Redshift, MySQL, Microsoft Fabric and other engines — with TLS, proxies and troubleshooting.
Importing from databases — BigQuery, SQL Server, PostgreSQL, Oracle, DuckDB and SQLite.
Importing files — Excel, CSV/TSV, JSON, NDJSON and Parquet imports, and the import history.
Importing from cloud storage — Google Cloud Storage, Azure Blob Storage and Amazon S3.

Export tab

Exporting to databases, files and cloud storage; multi-sheet workbooks; PDF reports; scheduled exports.
Export tab — controls — Every button on the Export tab.
Exports — destinations, formats and options — Paths and parameters, multi-sheet workbooks, scheduling, size limits and every export option.
PDF reports — Building a PDF report, page and branding options, security, destination and scheduling.

Functions tab

The formula bar, calculation modes and the complete function library.
Formula bar and calculation — Automatic and manual calculation, pausing a sheet, recalculating, and AI formula refresh.
Mathematical Functions — Syntax, arguments and examples for every function in the Mathematical category.
Statistical Functions — Syntax, arguments and examples for every function in the Statistical category.
Logical Functions — Syntax, arguments and examples for every function in the Logical category.
Text Functions — Syntax, arguments and examples for every function in the Text category.
Date & Time Functions — Syntax, arguments and examples for every function in the Date & Time category.
Lookup & Reference Functions — Syntax, arguments and examples for every function in the Lookup & Reference category.
Verification Functions — Syntax, arguments and examples for every function in the Verification category.
AI Functions — Syntax, arguments and examples for every function in the AI category.
Database (SQL) Functions — Syntax, arguments and examples for every function in the Database (SQL) category.

Reports tab

Pivots, project timelines, charts and shapes on a sheet; heatmaps, waterfalls, histograms, clusters and key drivers.
Reports tab — controls — Every button on the Reports tab, and how report objects behave on a sheet.
Project timeline — Milestones and phases, binding a timeline to a table, and its styling.
Shapes — annotating a report — Lines, arrows, boxes and speech bubbles; styling, anchoring to cells and drawing order.
Heatmap — A grid of cells coloured by value; date grouping, colour scales, rules, totals and outliers.
Waterfall — The bridge from an opening figure to a closing one; totals, subtotals and breakdowns.
Histogram — The distribution of one column: bins, range, outliers, curves, specification limits and statistics.
Clusters — segmentation — k-means segments found in several numeric columns at once: features, label, automatic k, scaling, the scatter and the profile, quality, reproducibility and cross-filtering.
Key drivers — What explains a number or a yes/no outcome: the Shapley ranking of drivers, direction and significance, the segments that stand out, R², the method and cross-filtering.
Pivot tables — Cross-tabs on a sheet or a dashboard.

Dashboard tab

Building dashboards, every widget and its properties, filters and cross-filtering, forecasting, sharing.
Dashboard tab — controls — Every button on the Dashboard tab, and how to build a dashboard.
Widgets and their properties — Title, KPI, filters, search box, pivot, table and chart widgets (including clusters and key drivers); binding, formatting and cross-filtering.
Range filter — Slicing a dashboard by a numeric range.
Forecasting — methods and parameters — Projecting a value over time: every forecasting method, seasonality, prediction intervals, scenarios, accuracy measures and every setting.

Ask Greg tab

The AI analyst: asking questions, grounding, model providers, data policy, audit and placing proposed objects.
Ask Greg — controls and governance — Every button on the Ask Greg tab and what the model may see and do.
Providers and models — The six model adapters, endpoint classes, the default model, the per-session model picker, where keys are kept, and models installed by your organisation.
Data policy and privacy — What may leave the machine: data levels and per-class ceilings, the payload preview, PII masking and its limits, saved connections, and what is and is not sent.
What Ask Greg can make — Every kind of object Ask Greg can propose, how each proposal is checked and graded (gold, silver, bronze), and how you place or remove it.
Scope rules and audit — The agent's fixed scope rules and how to inspect them, administrator rules, the SQL guards that actually enforce the boundary, the AI audit log and the licence components.
AI in formulas and pipelines — How AI_PROMPT and the pipeline AI analysis step relate to Ask Greg: the answer cache, the call cap, holding a sheet for review, and the boundaries they share.
AI compliance — EU AI Act and transparency — How the AI features map to the EU AI Act and US state rules, what the product does about each obligation, and what stays with your organisation.

Administration & security

What is protected and how: credentials, data at rest and in transit, project trust, object-level access, row- and column-level security, roles, audit, administrator policy, email and supply chain.
Security overview and data protection — What is protected and what is not; the credential vault; data at rest and in transit.
Application boundary and desktop shell — The backend's own boundary, the hardened desktop application, and opening projects saved by another version.
Project trust — opening projects from someone else — Why a project from another computer or person opens paused, what is paused, how to trust it, and how administrators pre-trust folders or switch the prompt off.
Roles, audit trail and logs — Server-mode roles, the tamper-evident audit trail, logs, files and support bundles.
Content catalog, certification and usage — One catalog of every sheet, dashboard, report, pipeline and query, with lineage; certification with a four-eyes review that lapses when the object changes; and a usage report that can show people by name, by pseudonym or not at all.
Security baselines, MFA and compliance evidence — Requiring multi-factor authentication at sign-in, the security baselines an administrator can enforce for a whole estate, and how the product's controls map to SOC 2 and ISO/IEC 27001 — with no certification claimed.
Object-level access and sharing — Per-object permissions on a server: share one project folder, dashboard, database or storage connection or mail profile with named people and Entra ID groups.
Row- and column-level security (RLS) — Per-user row filters and column masking for DuckDB tables, sheets and exports — enforced on the server for every query, with preview, explain and audit.
Administrator policy — Group Policy, Intune and the network — Central configuration, network controls and policy-enforced settings.
Email — mailbox and what leaves the machine — Mailbox reading and sending, and the controls that stop data leaving.
Sync and conflicts — Offline-first sync between devices and resolving conflicting edits.
Releases, updates and reporting a vulnerability — Signed releases, licences, updates and responsible disclosure.
Licences and activation — The signed licence file, the bundled evaluation, the Free edition, where the licence lives, installing one, the grace period and licensed modules.

REST API and MCP

The public REST API, headless metrics and MCP servers (BigQuery, PostgreSQL, Jira, Confluence and more): credentials, endpoints, conventions and errors, templates, security and resilience, Ask Greg with MCP, and operating it all.
API and MCP at a glance — What the public REST API, headless metrics and MCP servers are for, who uses which, the one security model under all three, and where to start.
The REST API — Credentials (scoped service tokens, Entra ID, the installation token), the role each operation needs, every endpoint of /api/v1, and worked curl and Python examples.
Conventions and errors — Cursor paging, JSON, CSV and NDJSON, rate limits, Idempotency-Key, load shedding, request ids, and every error code with whether to retry it.
Headless metrics — The semantic layer as a service: list the metric catalogue, see what a measure can be sliced by, and query measures by dimensions with governed filters and typed parameters — the same numbers the dashboards show.
MCP servers — Register BigQuery, PostgreSQL, Jira, Confluence or any other Model Context Protocol server from a template, test it, inspect and narrow its tools, and manage it over HTTP.
MCP parameters and settings — Every value you fill in for an MCP server — per template, with what is required, defaults, examples and meaning — how the Environment, Secret and Header boxes work, connecting BigQuery with a service account, and every installation setting for the API and MCP.
MCP security and resilience — How tools are classified and refused, the SQL guard, secrets in the vault, process isolation, egress rules, the machine policy switches, and the circuit breakers, bulkheads and deadlines that keep a failing server from taking the application down.
Ask Greg with MCP servers — The two switches that let Ask Greg use MCP tools, what the model is told and shown, how every result passes the data policy, and example questions across the project, BigQuery, Jira and Confluence.
Operating the API and MCP — Every server setting for the API and MCP, network exposure, audit records, Prometheus metrics, and what each common error message means.

Governed metrics, watches and embedding

The semantic layer in every surface — governed sources, METRIC(), the publish gate, compile checks and metric tests, personal definitions, time-window metrics, joins across connections, dbt and OSI import and export, the Governed metrics pipeline step and Ask Greg building on metrics — plus data stories and other AI narratives, insight watches, ATX AnalyticsTable as an MCP server, embedded dashboards and how Ask Greg's quality is evidenced.
Governed analytics at a glance — One definition of every number, used everywhere: in sheets, dashboards, cells, alerts, outside AI agents and other applications — and which of these features you need for which job.
Governed metrics in charts, pivots and dashboards — Point a sheet chart, a pivot table or a dashboard widget at governed metrics instead of a table or range — on the built-in database or on any warehouse connection — and keep its numbers identical to every other report.
The METRIC() formula — A governed number in one cell: =METRIC("revenue"), =METRIC("revenue", "region", "EMEA"), cell references as values, and arithmetic around it — evaluated on the server in one batch.
Publish gate and Restore — Serve reports, agents, METRIC() and embedded dashboards only the version of each metric a steward certified, and roll a definition back to any earlier saved version in one click.
Compile check and metric tests — Before a definition is certified, the compile check proves it runs on its database without reading a row; metric tests assert that its NUMBERS are right — six test types with examples, running and scheduling them, the badge, and the switches that make either block certification.
Personal definitions and Promote — Keep draft measures and what-if calculations that only you can see, use them wherever you would use a governed metric, and Promote them to the project or the shared catalog when they are ready — who has a personal scope, the rules, and the limits.
Time-window metrics — Running totals, year-to-date and month-to-date figures, rolling windows and period-over-period comparisons defined ONCE as governed calculated members — the fields, worked examples, the rules for periods, and the limits.
Joining across connections and fan-out correction — A governed metric can now combine datasets on different databases, and the planner corrects the double counting a to-many join causes — what you see, how each works, and the limits: a per-side row cap, and filters that must name their dataset. URN dashboards and embedded dashboards use the same planner.
Import and export: dbt MetricFlow and OSI — Bring dbt MetricFlow semantic models and metrics, or an Apache Ossie (OSI) model, into the semantic layer, and export your definitions to dbt YAML or OSI — what maps to what, what is left out and why, the safety checks, the lossless OSI round trip and a worked example.
The Governed metrics pipeline step — Read certified measures by dimensions into a pipeline instead of writing SQL, so a scheduled recipe gets exactly the figure the dashboards show — with a fail-safe for cut results, an optional certified-only rule, and a run log that names every definition and version.
Ask Greg building on governed metrics — Ask Greg does not only answer from governed metrics, it builds on them: charts, pivots and KPIs bound to a metric, pipelines with a Governed metrics step, and data stories — every reference planned against the catalog before you see a card.
Data stories — A short written narrative of a dashboard, table measures or governed metrics, in seven languages, on project tables or warehouse tables — every figure computed by the analysis tools, every sentence citing its sources, and every figure checked by code before you see it.
AI explanation boxes in PDF reports — A paragraph in a PDF report that explains one table, chart, pivot, KPI or dashboard of that report from its own numbers — figures checked against them, marked as AI-written, and replaced by a short note if the AI is unavailable.
AI paragraphs in alerts and Check with AI — An alert e-mail that explains itself in a few words, and a dry run that says whether a rule would fire on the current data and why — the verdict always decided by code, never by the model.
Insight watches (the watcher agent) — Alerts that watch a table or a governed metric for an anomaly in the latest period, a move against an earlier period, or a forecast crossing a line — with worked examples for each, and how the e-mail is written.
ATX AnalyticsTable as an MCP server — Let Claude, Copilot Studio, Cursor or your own agent ask ATX AnalyticsTable for governed numbers over the Model Context Protocol: switching it on, connecting a client with a service token or with OAuth sign-in (built-in or your own identity provider), the four tools, worked JSON-RPC examples, administering OAuth clients and the limits.
Embedded analytics — Show an ATX AnalyticsTable dashboard inside your own web application: configuration, minting a signed embed token on your backend, the JavaScript SDK, every widget type, filters, drill-down and Show rows, CSV/PNG export, theming, per-customer row scoping with locked parameters, revocation, audit and troubleshooting.
How Ask Greg's quality is evidenced — What the live capability matrix tests, the latest result, how numbers are re-checked against the database, what Ask Greg refuses on purpose, and how to run the matrix against your own model.

Reference

Keyboard shortcuts and the grid and sheet-tab menus.
Keyboard shortcuts — Every keyboard shortcut, by group.
Grid right-click menu — Commands on the grid's context menu, including converting pins.
Troubleshooting and common questions — Known failure modes with their cause and fix: installation, start-up, connections, server access, projects, networks, licences and where to find logs.

Glossary terms on this page

ATX Institute on LinkedIn

This site needs JavaScript for the full experience. Every page above is reachable without it.