Skip to main content

Search

Let us guide you!

Ask your dedicated IA server

Updated on

The Ask AI action with the Dedicated AI Server source lets you connect your automations to an AI server you fully control. It is the most sovereign and customizable option : your models, on your own infrastructure.

Full infrastructure control

Your models, on your infrastructure → hosted by TimeTonic or on-premise. You retain complete control over the deployed models, the data processed, and the processing capacity.

No TimeTonic credits consumed

This action does not draw from your monthly quota. No credit-related alerts or suspensions → billing is entirely separate from your other AI sources.

Performance tied to your server

Response time, processing capacity, and output quality depend directly on your infrastructure and the deployed model. If the server is unreachable, the automation will fail.

Available on the Business plan

This AI source is not available on the FREE, START, or PRO plans. Contact TimeTonic to set it up.

Overview

The dedicated AI server can be of two types:

Hosted by TimeTonic

Dedicated infrastructure managed by TimeTonic. Your models are deployed in an isolated environment, operated on your behalf.

Hosted by your organization

On-premise or private cloud. You connect TimeTonic to your own AI server, accessible via a URL and authentication credentials.

Prerequisites

Before using this action, an AI server must be configured and reachable. You will need the following:

  • An access URL for the server
  • Valid authentication credentials if the server requires them
  • One or more models deployed and running on that server

Server configuration is typically handled by a technical administrator. If your server has not yet been set up, contact your IT team or reach out to TimeTonic.

Model detection

When configuring the action, TimeTonic automatically queries the server to detect available models. They then appear in the selection list.

If no models appear in the list, check the following:

Server accessibility → Is the URL correct and reachable?
Model configuration → Are the models properly deployed and active on the server?
Access permissions → Are the authentication credentials valid?

Configuring the role and prompt

As with all AI actions in TimeTonic, two fields structure your call to the model:

Role description

Defines the expected behavior and context for the model.
Example: "You are an assistant specialized in financial analysis."

Prompt

The precise instruction sent to the model. Use $ to dynamically inject values from your fields.
Example: "Extract the following information in JSON format."

💡 A clear role and a precise prompt significantly improve the quality and consistency of results produced by the model.

For PDF or image analysis, a specific prompt is required → See OCR PDF & images prompts

Attachments

Select a field containing image files to pass to the model. These files will be included in the request if the model installed on your server supports document analysis.

📄 Document analysis
🔍 Information extraction from files
🖼️ Image processing

⚠️ Attachment support depends on the model deployed on your server. Verify this capability before using it in production.

For PDF file processing → See the OCR PDF & images procedure

Token parameter

The Maximum token count field caps the length of the response generated by the model.

Token limit Effect on response Use case Recommended value
Low
Short, concise response: the AI gets straight to the point Classification, single-value extraction, category, yes/no 50 to 200 tokens
Medium
Structured response in a few lines Suggested email reply, short reformulation 100 to 400 tokens
High
Long, detailed response: summary, report, full analysis Document summary, meeting notes, content generation 500 to 1,000 tokens
Adjust this parameter to match the nature of the task → limiting tokens helps control your TimeTonic credit consumption. A 100-token response costs ten times less than a 1,000-token response.

The token count directly impacts your credit consumption. To anticipate the impact before configuring your action → Estimate your AI credit consumption

6. Response format

The response format determines how the AI-generated response will be structured and stored in your notebook. Two options are available:

Text

The response is stored as-is in a text field. Ideal for summaries, comments, or free-form responses.

Structured JSON

Each JSON property is stored in a dedicated field. Recommended for structured extractions and automated processing.

💡 The Structured JSON format requires a specific prompt in the Prompt field to instruct the model on which data to extract and in what structure → especially for PDF or image analysis. Each JSON key then maps to a destination field in your record.

See the OCR PDF & images configuration →

7. Response field

This field determines where the AI response will be stored in your notebook. Its behavior depends directly on the format selected in the previous step.

If you selected the Text format

The Response field appears. Select the text field in your notebook where the generated response will be stored. Only one destination field is required.

If you selected the Structured JSON format

The Response field disappears and is replaced by a mapping table. For each JSON property you requested in your prompt, you must:

  1. Enter the exact name of the JSON property as you defined it in your prompt (e.g., name, date_of_birth, amount)
  2. Select the corresponding destination field in your notebook where the extracted value will be stored

Repeat for each property to extract. You can add as many rows as needed using the + button.

Example: You are extracting data from an ID card and your prompt requests a JSON with the keys last_name, first_name, and date_of_birth. You create three rows and map each key to the corresponding field in your notebook.

Best practices

  • Verify attachment compatibility before using this option in production → not all models support it.
  • Write a precise and consistent role aligned with the expected task → this is the single most impactful factor on output quality.
  • Limit the token count to match the actual requirement and avoid unnecessary load on your server.
  • Test the action on a real use case before deploying at scale → validate output quality and connection stability.

Learn more

AI credits by license plan

How credits work, available sources, and the 1,000-token rule.

Read the article →

Learn more

Use your own API key (BYOK)

Connect your provider subscription directly to your automations.

Read the article →

Learn more

View consumption in the logs

Analyze the details of each AI call and its consumption per execution.

Read the article →