You’ve created your account. Now what? This guide walks you through the platform end to end, entirely in the web app — no code editor, API, or SDK required. By the end you’ll have connected data, created an ontology, run your first concept, explored the results, and built a app on top of them.
This is the app walkthrough. Once you’re comfortable clicking around and want to learn the Vadalog language itself, continue with Your First Ontology and the Vadalog Reference.

​
Step 1 — Get your bearings

After signing in and completing the “Complete your profile” screen, you land in your workspace. The home screen centers on the AI Assistant, and the left sidebar is how you move around:
A Getting Started checklist lives at the bottom of the sidebar and tracks your progress through the first few tasks (connect data, create an ontology, create an app, and more). It’s a handy companion to this guide.
Prometheux workspace home

The workspace on first login — AI Assistant in the center, navigation sidebar on the left, and the Getting Started checklist at the bottom of the sidebar.

​
Step 2 — Connect your data

Open Data (/datasources) from the sidebar. This page has two tabs: Sources (database connections) and Files (uploads).
1

Open the connection dialog

Click Add Data Source. The “Add your data source” dialog opens with a searchable, categorized list of connectors — Relational Databases (PostgreSQL, MySQL, SQL Server, Oracle, …), Cloud Services (Snowflake, Databricks, BigQuery, Redshift, …), NoSQL (MongoDB, Neo4j), Vector Stores, and File Formats (CSV, JSON, Parquet, Excel, …).
2

Pick a connector and fill in the details

Select your source type and complete the connection form. Fields vary by type — for example, Snowflake asks for Username, Password, Account URL, Database, and an optional Warehouse; PostgreSQL asks for Host, Port, Database, Username, and Password.
3

No data handy? Use a demo source

On a fresh workspace, expand “Available px-datasources — no setup needed” and pick a read-only Prometheux dataset. It opens the connection form pre-filled — just review and connect, so you can explore the platform before wiring up your own systems.
4

Or upload a file

Switch to the Files tab, click Upload, and choose a CSV, Excel, or other supported file. Each uploaded file gets a Connect button that registers it as a data source. (Excel workbooks let you pick which sheets to connect.)
Once connected, your source appears in the Sources list with its record and column (feature) counts. Use Preview to inspect the data in a paginated, searchable table, and Run to use it inside an ontology.
Add data source dialog

The 'Add your data source' dialog — categorized connectors with a search box, and the 'Available px-datasources' demo section expanded.

​
Step 3 — Create an ontology

An ontology is your workspace for turning data into answers — it holds your concepts, schema, and apps. Open Ontologies (/ontologies). On a new account you’ll see “It looks like you don’t have any ontologies yet.”
1

Start a new ontology

Click New Ontology (top-right). In the dialog, stay on the Create tab and fill in:
  • Name — e.g. Company Ownership
  • Description (optional)
  • Context (optional) — a few notes about the ontology’s goal; the AI Assistant reads this on every turn.
  • Scope — leave as User Scope.
2

Create it

Click Create. You’re taken straight into the ontology editor, opened on the Lineage view.
In a hurry? If a Templates tab is available, hover a template and click Use → to spin up a ready-made ontology (schema, concepts, and sample data included) that you can explore and adapt.
Create a new ontology

The Ontologies page empty state with the New Ontology button (top-right), and the New Ontology dialog showing the Name, Description, Context, and Scope fields.

​
Step 4 — Define your first concept

Inside the ontology editor, the Explorer panel on the left lists your concepts — the logic that transforms data into results.
1

Add a concept

Click New cell in the Explorer header. A new concept cell appears.
2

Choose how to express it

Pick a concept type: Logic (Vadalog, the default), SQL, Cypher, or Python. Write your logic in the editor.
3

Let the assistant write it for you (optional)

Type a plain-English description of what you want, then click the ✨ Generate Vadalog from natural language button. Prometheux drafts the Vadalog for you; edit it as needed.
4

Save

Save with the Save cell button or ⌘S / Ctrl+S. An unsaved cell shows a Modified badge.
New to Vadalog? The Your First Ontology tutorial teaches the language step by step, and the Vadalog Reference is the full guide.
Defining a concept

The ontology editor — the Explorer panel with a concept cell open, the concept-type selector (Logic/SQL/Cypher/Python), the CodeMirror editor, and the ✨ natural-language button.

​
Step 5 — Run it and explore the results

With a concept saved, run it from the cell toolbar:
  • Run cell executes just the current concept.
  • Run workflow runs the concept together with everything it depends on.
Press Shift+Enter as a shortcut. While it runs, the button becomes a Stop control and the results panel shows “Running query…”. When it finishes, the Output tab opens automatically with your results. In the Output table you can:
  • Page through rows (Prev / Next, with a Rows size selector).
  • Search across columns and apply per-column filters (contains, equals, <, >, …).
Switch to the Lineage view (open the Explorer rail) to see your concepts as a dependency graph — click any node to jump into its editor.
Exploring concept results and lineage

A concept after running — the Output tab showing the results table with search and pagination, and the Lineage dependency graph alongside.

​
Step 6 — Build an app

Now turn your results into something shareable.
1

Create an app

Open Apps (/apps) and click New App.
2

Point it at your ontology

In the config panel, choose your Ontology and an Input concept (drives run) — the concept whose run produces the data your widgets display. Set a Title and any header details.
3

Lay out pages, sections, and widgets

Click Add page for a tab, Add section within it (choose Single, Two columns, or Four columns), then Add widget. Pick a widget type — Table, Line chart, Pie chart, or Progress bar — and select the Concept it visualizes. For charts, map the X axis / Y axis (or label / value) columns.
4

Save and run

Click Save, then open the app and click RUN. Prometheux executes the input concept and populates every widget with live data.
Building an app

The app editor — the config panel (Ontology, Input concept, pages/sections) on one side and the widget canvas with a chart and table on the other.

​
What’s next

You’ve gone from an empty account to a running ontology and a live app. From here:

Learn Vadalog

Write your own reasoning rules with the step-by-step language tutorial.

Connect more data

Wire up Snowflake, Databricks, SQL databases, and files.

Use the AI Assistant

Have the assistant generate and debug concepts for you.

Automate & integrate

Drive the platform from the REST API, Python SDK, or MCP.