Squirro Introduces a RAG-Powered Digital Assistant to Redefine Online User Engagement

Zürich, December 7th, 2023. Squirro, the generative AI-enabled semantic enterprise search provider, is proud to announce the launch of SquirroGPT Web, one of the first Retrieval-Augmented Generation (RAG) digital assistants.

This cutting-edge generative AI application is set to redefine how organizations engage with their online audiences. It provides dynamic, real-time interaction with website visitors and conversationally guides them to the most relevant content and resources.  

SquirroGPT Web sets itself apart from standard chatbots or digital assistants thanks to its unique technology. Leveraging a robust Information Retrieval (IR) stack, this application seamlessly integrates high-quality semantic search to fetch accurate and relevant information. Additionally, it employs a potent RAG model, ensuring high-quality responses. These responses are not only accurate but also transparent, as verifiable sources back them.

Squirro has more than ten years of expertise in data lifecycle management and has developed a robust IR stack leveraged by world-leading companies across industries.

“Thanks to our background and experience in data science and engineering, we’ve developed a unique and innovative digital assistant for websites, which combines the advantages of LLMs in terms of accessibility with the data accuracy and security that our proven and robust IR stack provides,” said Dr.  Dorian Selz, CEO and Co-founder of Squirro.

Users can try SquirroGPT Web on websites such as docs.squirro.com and squirro.com.

An Innovative Digital Assistant

SquirroGPT Web crawls data from websites and enterprise documents, including user manuals. This enrichment enhances information access, significantly increases user engagement, and boosts the ROI of any website by implementing a high-quality digital assistant.

This application is part of the SquirroGPT offering, with differentiated features and benefits:

• Specific and transparent answers based on your company data. Squirro's retrieval-augmented generation (RAG) guarantees that every result is rooted in evidence, offering direct links to the source and ensuring informed decisions based on trustworthy data. In that sense, the user can cross-reference the source of the answer in just one click.
  
  
• Access control to protect confidential information. The provided answers are compliant with each user’s access levels.
  
  
• Excellent white-label capabilities to make the integration of SquirroGPT Web on any website from a branding perspective easier.
  
  
• Interaction feedback opportunities for users to evaluate the provided answers. This is key to improving the quality of the digital assistant over time, learning from the users’ feedback.
  
  
• Multilingual UI support allows users to choose their desired language for the application UI.
  
  
• Language specification gives users the possibility to ask questions in their preferred language regardless of the UI language setting.