At #Acelera2023, Athento’s first Open Document Management Conference, our colleagues showcased several AI applications to content management that were already available for use on the platform.
Below, we bring you the latest news about Athento and the most important AI applications that were showcased at the event and share information about some of the latest developments in the field.
Hugging Face for AI operations
First, our colleague Iván Cárdenas (min. 7:47 approx.) explained the use of the integration with Hugging Face for AI operations. He introduced how this leading platform in the field of Artificial Intelligence works, which has the support of many researchers and which, in addition to having many artificial intelligence models, focuses on natural language processing models (same neural network architecture as Chat GPT, Google’s BERT), pre-trained with a large amount of data.
As we saw, Hugging Face has an open source philosophy, so it has a large number of models, datasets and totally free tools that open up a very broad horizon of possibilities to apply it to many types of tasks such as text processing, image processing and even sound processing. It is also a technology trained in multiple languages.
And all the power, tools and possibilities of Hugging Face are available in Athento’s operations.
As a specific demonstration of AI applications that can already be used by our customers, our colleague showed the use of a sentiment analysis model in a transaction. That is, a model that from text processing is able to know what the emotional tone of the user writing the text is. This is helpful for companies to know if a product, service or idea they have is liked or disliked by customers by knowing directly how they feel.
Chat GPT as a wizard for creating regular expressions in data extraction templates
The next functionality related to artificial intelligence that we showed, in this case, by our colleague Alberto Parras (min. 25:30 approx.), was the ‘Definition of regular expressions using Chat GPT’.
This functionality is especially useful for non-technical users. This specific application of Athento’s integration with Chat GPT allows to generate regular expressions to create the data extraction templates of a document.
In the video, you can see a live Demo on how to generate a regular expression to extract a DNI from the text of a document, in a matter of seconds!
In addition, our colleague discussed future applications of AI to use Chat GPT and Athento, such as:
- Developing programming code for operations.
- Inquiring about the difference between the content of documents or different versions of the same document.
- Queries about the content of a document.
- Support for the configuration of a document type or form.
DialogFlow and Athento integration for creating custom chatbots
Finally, our colleague Adriana León (min. 33:15 approx.) explains the integration of Athento with Google’s Dialogflow tool, which allows the creation of conversational artificial intelligences or chatbots. It is possible to configure these chatbots to carry out certain actions in the document manager through a conversation.
Our colleague Adriana explained what Dialogflow is and how it works, a Google service that allows you to create chatbots or virtual agents.
As a specific demonstration, Adriana showed a live demo of a pre-configured chatbot for a real use case: users applying for jobs from the Athento website through a conversation with the virtual agent. This is one of the applications of AI transversal to any company, which can be activated today and which allows homogenising and automating tasks in recruitment and selection processes.
In addition, the general concepts of how Dialogflow works and how to configure the chatbot to connect it with Athento were explained. On the one hand, a tool to provide a graphical interface to the chatbot was shown. And on the other hand, how customisable the integration is.
More AI applications you can start using today with Athento
These AI applications join other AI functionalities already offered by Athento such as:
- Predicting metadata in documents.
- Generating graphs with information about documents.
- Prediction of the next state in a life cycle.
- Classification of emails.
And, recently, we have launched a Digital Transformation Recipe or CookBook in which we explain step by step how to apply AI to facilitate searches in digital image repositories. AI applied to image searches can upgrade retrieval times and reduce manual indexing or cataloguing work.
What we share in the video and in the documentation is the current state of these technologies in Athento, which are already a reality at production level. However, the applications of AI to document and business content management have a very broad horizon ahead in which the only limit we have is our imagination.
If you want to boost your business processes related to documents and content, write to us and tell us about your challenges.