T3AC 2.0: AI Chatbot for Better Control, Visibility, and Workflow

T3AC 2.0 provides a clear and controlled way to manage how data is added and updated in the system.

Supported Data Sources Across TYPO3 and External Systems

The chatbot can connect to multiple data sources:

• TYPO3 pages and content
• News extension
• Web pages and sitemap XML
• PDFs and text documents
• Q&A pairs
• Apache Solr, ke_search, indexed_search
• API-based sources (planned)

Per-Source Sync Control and Queue Visibility

Each data source can be managed independently:

• Enable or disable specific sources
• Set sync frequency per source
• Trigger manual “sync now” actions
• View sync queue and processing status per source

Training Center - A Transparent AI Pipeline

Earlier, there was no clear way to see what was happening during training. Now, every step in the process is visible and manageable.

The Training Center provides a clear view of how data is processed and prepared for the chatbot.

Dedicated Training Interface Inside TYPO3

A structured interface to manage the full training pipeline:

• Central place to manage all training data
• Clear separation of data processing stages
• Better visibility into what the system is learning

Status Tracking and Error Handling

Each item in the pipeline can be tracked and managed:

• Status levels: pending, fetching, embedding, completed, failed
• Re-queue failed items for processing
• Remove stuck or unnecessary entries
• Token visibility per item for detailed usage tracking

Central Dashboard - Full Visibility in one place

The central dashboard is a completely new addition in T3AC 2.0, giving a clear overview of how the AI system is being used.

Token Usage, Costs, and Activity Monitoring

Track system usage and performance in one place:

• Token and embeddings usage tracking
• Estimated cost visibility
• Interaction and activity overview
• Status of training pipeline

Managing AI Operations from One Place

Control key actions directly from the dashboard:

• Overview of active modules
• Quick actions for common tasks
• Simplified monitoring for daily operations

Advanced AI Settings & Control Layer

T3AC 2.0 introduces deeper control over how the AI behaves and operates.

LLM Provider Selection and Response Tuning

Adjust how the chatbot generates responses:

• Select and manage AI model providers
• Control temperature and response style
• Set maximum token limits

Budget Management and Cost Control

Keep AI usage within defined limits:

• Set monthly usage budgets
• Configure budget alerts
• Apply hard limits to control spending
• Especially important for larger projects with ongoing AI usage

RAG Configuration for Better Context Handling

Improve response quality with structured context control:

• Define context documents
• Adjust chunk size and overlap
• Optimize how data is used for responses

Flexible Model Support for Future Use Cases

Support for evolving AI requirements:

• Works with multiple LLM providers
• Easy switching between supported models
• Designed for long-term flexibility and updates
• No system changes required when models evolve

CLI Scheduler for Developers

The CLI scheduler adds more control for development and automation tasks.

Automating AI Processes via Command Line

Run and manage processes without manual steps:

• Execute tasks via CLI
• Schedule automated processing
• Handle large data operations efficiently
• Dry run option to test processes without execution
• Process tasks by specific source ID
• Better control for staged testing and debugging

Debugging, Reprocessing, and Batch Control

More control for testing and maintenance:

• Verbose and debug modes
• Reprocess specific data sources
• Batch size control for operations
• Cleanup and re-queue options

Chatbot Customization & User Experience

T3AC 2.0 provides more flexibility in how the chatbot looks and behaves.

Controlling Chatbot Behavior and Tone

Define how the chatbot interacts with users:

• Set instructions for tone and responses
• Configure welcome and system messages
• Enable or disable chatbot globally
• Select language preferences

UI Customization and Frontend Experience

Adjust the chatbot interface to match your website:

• Customize avatar (including sizing)
• Show or hide logo
• Set bubble message and welcome message
• Transparent trigger option
• Control position (left/right) and bottom spacing
• Theme colors and UI elements (date, time, etc.)

Flexible Embedding & Deployment Options

The chatbot can now be deployed in more flexible ways.

TYPO3 Native Integration

Seamless integration within TYPO3 frontend:

• Automatic embedding in TYPO3 pages
• Works within existing site structure

External Embedding Possibilities

Extend chatbot usage beyond TYPO3:

• Embed on external websites
• Flexible deployment across platforms
• Use in different environments as needed

Usage Analytics & Insights

T3AC 2.0 provides full analytics with filters, tracking, and export options. It     adds clear insights into how the chatbot is being used.

Tracking Interactions, Sessions, and Feedback

Understand user behavior and system usage:

• Total interactions and session tracking
• Query-level insights
• Positive and negative feedback tracking
• Full conversation logs (Q&A)

From Chat History to Actionable Insights

Move beyond basic logs to meaningful analysis:

• Filter data by module or language
• Identify common queries and gaps
• Export reports for further analysis

Conclusion

AI Chatbot 2.0 brings three key improvements to TYPO3 projects: clear visibility into how AI works, better control over data and costs, and a structured system that fits into existing workflows. 

Instead of treating AI as a separate layer, it becomes part of the CMS with defined processes and manageable outputs.

This shift also reflects a more practical approach to AI. The focus is on control, automation, and insight, supporting teams in handling repetitive tasks while keeping decisions and responsibility where they belong. Download free!