A New Chapter: AI-Powered Document Management Experiences with Uxopian AI

Last year, at UxoDays in the beautiful Haussmann-style room of the Cercle National des Armées in Paris, we announced our mission: to bring the right information to the right place at the right time, securely. We also shared our intention to use AI as a means to achieve this.

One year later, we are proud to share the first results of these efforts. Uxopian Software solutions now feature AI capabilities designed to reduce the time needed to complete tasks, while improving the quality and consistency of the outcomes.

There are a few pieces of good news:

• These capabilities are available today for customers who have migrated to ARender 2023.X and wish to join our early adopters program.
  
  
• They will be fully integrated into our major product releases in early 2026.
  
  
• We have also created a framework, currently called the Uxopian AI Framework, that we use internally to integrate AI capabilities into our products (Fast2, ARender, and Flowerdocs). Because it has proven so useful, we are now making it available as a standalone product for organizations and software vendors.

Watch the introduction to the Uxopian AI Framework

×

The Uxopian AI Framework helps accelerate the adoption of LLM-based technologies and unlock their business potential more quickly, no matter your infrastructure choices: self-managed, on-premise, our cloud, or hybrid setups, and with any AI provider.

What does this mean in practice?

ARender, Flowerdocs, and Fast2 can now include an optional assistant that understands the context of the document being opened or processed. This assistant can be used for:

• free conversation with context,
• pre-configured prompts for fast and reliable information extraction,
• or even as a natural-language interface for performing content management actions (create, move, archive, export, etc.).

This “surface” integration of Uxopian AI can be applied not only to our own products, but to others as well. From the start, we designed it to be extensible: if you are using ECM X, for example, you can contribute your own prompt helpers to fetch content, manage taxonomy, and more. It could even be connected to your CRM or ERP.

In the coming quarters, we will go further, embedding AI more deeply into the core user experience of ARender and Flowerdocs, with Uxopian AI quietly powering features behind the scenes. But even with today’s assistant-based user experience, the value created is already significant and worth sharing.

Discover “Zero Scroll” and see how end users can benefit from these new AI-powered experiences

×

Here are a few use cases available right now:

• Automated information extraction (e.g., completing metadata at ingestion time)
• Automated content validation (e.g., checking documents against a policy and flagging what is missing or non-compliant)
• Summarization
• Translation
• Comparison

All interactions are tracked: we provide per-use-case token consumption, usage analytics, and user feedback, so organizations can monitor the impact of their AI initiatives. The framework is compatible with most major AI providers (OpenAI, Google, Anthropic, Mistral, Hugging Face open models) and can also work with your in-house APIs through a simple Java class implementation.

We also designed our bot component as a reusable web element to simplify integration. Importantly, it is not just a “proxy” to an LLM: thanks to function and tool calling, the assistant has direct access to ARender and Flowerdocs functionalities.

This means many actions that once took multiple clicks and time can now be completed instantly. We believe this opens the door to re-inventing the user experience of content management. And if you want to extend Uxopian AI for another software project, you can easily expand the toolset.

This is only the beginning of what we see as a very exciting decade ahead. Imagine combining the orchestration power of Fast2 with intelligent AI-driven automation! What our team achieved in just one year gives us confidence in the road ahead, and we look forward to what’s coming next.

If you want to know more, experiment with these capabilities, or simply share your current challenges in applying AI to real business problems, we would be very happy to exchange with you.