Easy to understand.Complete enough for experts.
Your table data is not copied into ClarioDB’s system database: it passes through only to display, export or perform what you request. You understand what will happen, keep a recovery point and remain in control of the final click.
SQL Studio, ClarioDBScope, exports and Safe Preview work on the active database without creating a parallel warehouse. The 4i review, still in design, will be triggered only on request, with a reduced context containing no credentials, row values or results, and will not be able to execute anything.
The databases that matter.
1 UPDATE orders 2 SET status ='archived'3 WHERE created_at <'2024-01-01'; 4 5 -- understand, protect, execute
Nothing has been changed yet. The report and recovery point are prepared before execution.
A continuous workflow, not a collection of tabs.
Traditional tools expose database objects. ClarioDB organises the complete path from intention to action : recover the right context, understand, protect, execute, and be able to roll back.
Your context is waiting for you.
A connection is no longer a form to fill in again: it becomes a workspace you can return to.
Start in the right place, immediately.
ClarioDB remembers authorised connections, database context, and user preferences. You resume where you left off.
No need to recover the host, port, and right tool for every intervention.
The workspace keeps preferences, skins, and the path to technical views.
Being guided never means being limited.
ClarioDB does not create a simplified version and a “real” version. It presents the same capabilities with the level of explanation and density you choose.
Guided by Ada or Grace
For managing a website, store, or project without having to learn every DBMS detail before acting.
- Operations are explained in plain language.
- The workflow shows what will change before execution.
- The pre-operation backup is built into the workflow.
- Technical details remain available at all times.
Advanced mode for professionals
For developers, administrators, and agencies that want the terminal, schema, metrics, and logs in one continuous workspace.
- Syntax-highlighted SQL terminal and multi-statement scripts.
- DDL/DML classification, scope, and technical risk.
- ClarioDBScope as a panorama or ERD canvas.
- Exports, history, backups, and user preferences.
The current interface, without staging.
Every screenshot below comes from the currently shipped interface, using a demonstration connection and dataset. No conceptual mock-up stands in for the real product.
Everything that truly matters, organised around the decision.
ClarioDB keeps the features expected from a database tool, then adds the memory, context, and protection workflow that are often missing between them.
Connection vault
Remember, organise, and recover your access details and each database context.
Terminal SQL
Syntax highlighting, scripts, results, history, and exports in a dedicated view.
Safe Preview
A pre-execution report exposes scope, strategy, and recovery point.
Integrated backups
History, downloads, and a dedicated backup before complex operations.
ClarioDBScope
Relational panorama for fast reading; ERD canvas for exploring and moving tables.
Data QuickLook
Inspect a long value, binary content, or a row without losing your working position.
Three skins
Continuity, Workbench, and Atelier: visuals change while the engine and permissions remain identical.
Context-aware autocomplete
Tables, columns, aliases, functions, and relations suggested from the active schema.
Your databases stay yours.
ClarioDB does not pull in your data to build a parallel warehouse. Its system database keeps what runs your workspace — account, preferences, labels and technical metadata — not a continuous copy of your tables or results.
No silent replication
ClarioDB queries the active database while you work. Rows and results are not copied into its system database.
Protected secrets
With a direct connection, credentials are encrypted with your passphrase before storage. With the Agent, they remain in the file deployed on your server. Its vault prefers a private location outside the web root; when a host exposes none to PHP, ClarioDB accepts a hardened web vault only after externally proving that its canaries cannot be reached. Local private keys remain encrypted there as well.
Explicit backups
An archive exists only because you request it or a sensitive write requires a recovery point. It is encrypted and its retention period is disclosed.
No secrets or row data for AI
The future 4i review will receive no credentials, row values, results or backups. For structural work, it sees only a pseudonymised manifest with no real names or source SQL.
Encrypted manual backups remain available for seven days. The temporary copy created by Safe Preview is deleted from the server after your receipt acknowledgement.
A panel provides access. ClarioDB organises the work.
ClarioDB does not try to hide SQL or simplify the profession until it loses substance. It connects the steps users currently rebuild across several screens, tools, and habits.
Security must be designed, explained and owned.
Emavio designed ClarioDB, its architecture, interfaces and protection protocols. Infomaniak is the hosting partner chosen by Emavio — not the product publisher.
Emavio SRL remains responsible for the product.
ClarioDB is designed, developed, maintained, and operated by Emavio SRL. Security choices, code, workflow, and user relationships remain Emavio’s responsibility.
Infomaniak infrastructure
Emavio chose Infomaniak to host ClarioDB and run service-side processing on dedicated servers in Switzerland.
AI works with you.Never instead of you.
ClarioDB clearly separates local assistance, AI advice and execution authority. Moving from one level to another is visible, optional and initiated by you.
Ada and Grace explain the engine.
Their answers come from ClarioDB deterministic checks and Knowledge Layer. No generative model or external AI call is required.
The 4i review will add a second opinion.
The intended context will be shown before any consent. The opinion will remain optional, identified and non-executable.
ClarioDB verifies. You decide.
The engine checks every proposal against its deterministic rules. AI has no execution tool: only your explicit action can apply a supported operation.
Connection credentials, row values, query results and backups are excluded. For a structural plan, the manifest is pseudonymised: no real table or column names and no source SQL.
- No automatic call
- No execution rights
- No hidden decisions
- Provenance always displayed
About Euria: it was used as an exploratory second opinion while some protocols were being designed. It is neither ClarioDB’s engine, a decision authority nor a certification.
AI turned around to serve your decision.
IV, four in Roman numerals, turns into something close to AI: AI put back in the right direction, towards your decision. With the Agent, a structured change can now join a local batch without immediately touching the database. ClarioDB shows actual and projected state, checks conflicts and lets you decide when to apply the batch. The four-I review will follow as an optional second opinion.
Intentions
What the batch appears to be trying to achieve, restated without deciding for you.
Impacts
The consequences, dependencies and volumes that deserve attention before the final click.
Inconsistencies
Contradictory operations or sequences that may reveal an incomplete intention.
Uncertainties
What cannot be asserted before execution and must remain clearly identified as such.
Your data stays yours. So does the decision.
ClarioDB combines full SQL power, deterministic safeguards and assistance that explains without taking ownership of your database or acting in your place.
