Our technology

We employ a variety of AI algorithms, carefully selected based on their suitability for diverse applications. For straightforward Question & Answer scenarios, we've crafted a user-friendly semantic search algorithm. This algorithm stands out for its simplicity, requiring no training. Upon updating the data, the application becomes instantly operational, making it easy to operate even for individuals with basic computer skills. Furthermore, content modifications are effortlessly managed.

More sophisticated applications dealing requiring proprietary data, we leverage the latest Retrieval Augmented Generation (RAG) models, predominantly relying on OpenAI's LLMs. For clients equipped with powerful GPU cards, we offer several smaller models tailored specifically to their requested applications. This customization enhances accuracy and reduces latency. RAG is capable to answer complex questions from the supplied content. It also answers "I do not know" if the answer is not found in the supplied content.

Our expertise extends to applications specializing in statistical analytics. These models process large mainly numerical datasets, such as web application logs, using a natural language interface. Users can ask queries in a conversational manner, in natural language, eliminating the need for regular expressions or SQL queries. The application then generates tables or graphs, providing a comprehensive analytical view.

The graphic below visually outlines the NLP pipeline, a common architecture among all the aforementioned applications.

The Smart AI pipeline 

We present our SocialBot -- Alquist~5.0 -- developed for the Alexa Prize SocialBot Grand Challenge~5. Building upon previous versions of our system, we introduce the NRG Barista and outline several innovative approaches for integrating Barista into our SocialBot, improving the overall conversational experience. Additionally, we extend our SocialBot to support multimodal devices. This paper offers insights into the development of Alquist~5.0, which meets evolving user expectations while maintaining empathetic and knowledgeable conversational abilities across diverse topics.