AI in TimeTonic enables you to analyze, structure, summarize, or transform data automatically within an automation workflow.
It is used via a dedicated action: "Ask an AI".
This action can be configured with different AI sources depending on your requirements in terms of data sovereignty, billing, or infrastructure.
Available AI sources
When configuring an automation workflow, you select the source to use. Here are the three sources available in TimeTonic:
* When you create your account, a 30-day Trial plan gives you access to the features and credits of the PRO plan. Once the trial period ends, your account automatically switches to the FREE plan.
How to add an AI action in TimeTonic?
In an automation workflow, AI acts as a step in the scenario.
Basic flow:
- A trigger fires (e.g., a field is updated, a document is added).
- A condition can be evaluated (e.g., field is not empty).
- The "Ask an AI" action is executed.
- The response is written to one or more destination fields.
Selecting your AI action in the configurator
When you add a step to your automation scenario, the Elements panel displays the AI actions available based on your license plan.
💡 AI actions appear under the AI section of your automation configuration panel.
⚠️ If you do not see the Ask a Private & Secure AI action in your configurator, check your license plan. This action is available from the PRO plan onwards. The Ask your Dedicated AI Server action is reserved for the BUSINESS plan.
* When you create your account, a 30-day Trial plan gives you access to the features and credits of the PRO plan. Once the trial period ends, your account automatically switches to the FREE plan.
What changes depending on your AI source
Depending on the selected AI source, the upper section of the configuration panel changes. The lower section → role, prompt, attachments, token limit, response format, and destination field → remains identical in all cases.
The Use a personal API key toggle is disabled ⚪. Only the AI Source and Model fields are visible. The model is selected from those available via TimeTonic credits.
The Use a personal API key toggle is enabled 🔵. Two additional fields appear: Provider (ChatGPT or Mistral AI) and API Key to enter your key. The available models correspond to those offered by your selected provider.
Only the AI Source and Model fields are visible. No toggle, no API key. The model is selected from those available on TimeTonic's sovereign infrastructure.
Only the AI Source and Model fields are visible. The displayed model is auto-detected from your server → it may be a custom or open-source model deployed by your organization. No API key or external provider is required.
💡 The lower section of the panel → role, prompt, attachments, token limit, response format, destination field → remains identical regardless of the source. Only the source configuration at the top of the panel varies.
* When you create your account, a 30-day Trial plan gives you access to the features and credits of the PRO plan. Once the trial period ends, your account automatically switches to the FREE plan.
Fields common to all sources
Regardless of the selected AI source, every "Ask an AI" configuration panel includes the same core fields:
💡 The AI Role field is optional but significantly improves output quality.
Credits and consumption
AI actions in TimeTonic consume credits based on:
- The number of tokens sent (input) and generated (output)
- The selected model
Consumption is visible in your automation logs and your license plan counters..
AI actions using your own API key do not consume any TimeTonic credits. Billing is handled directly by the provider (OpenAI or Mistral).
Typical use cases
- Summarize a text
- Classify an incoming request
- Extract structured information
- Analyze a document (invoice, resume, contract)
- Generate a personalized response
- Clean or reformat data
Best practices
- Define a clear role for the AI
- Write a precise, unambiguous instruction
- Limit the token count when necessary
- Test with realistic data