Domo has officially announced the launch of several new features for its Domo AI framework, which is a collection of flexible AI services. Talk about these features on a slightly deeper level, we begin from AI Chat, a solution designed to give you the ability to hold contextual conversations with your data. Markedly enough, the solution arrives on the scene as capable of figuring out what data is most relevant depending on the dashboard or app a user is viewing. Furthermore, the AI Chat even preaches optimal transparency at every touchpoint in your operations, thus allowing a user to see the exact steps taken when answering a question, while simultaneously making it possible for them to edit the SQL used in its answer. Not just that, it also helps users in building faster dashboards and generate insights, all done through a conversational way for smarter and faster decision making. Next up, we have a set of universal models that offer high code value for low code users. Here, you have out-of-the-box AI models that are accessible to everyone, as well as for use cases of all complexities. Domo is expected to launch Universal Models for forecasting next month using its AI Services Layer. As for other ones, models for PII, sentiment analysis, and anomaly detection will become generally available in the coming months, and once they do, they will be accessible from apps, cards, workflows, Magic ETL, and other avenues.
“Our goal with Domo.AI is to provide customers with an ideal environment to merge their invaluable data with the transformative power of AI in order to supercharge business results,” said Daren Thayne, chief technology officer and EVP of product at Domo. “By creating that type of open framework for AI + Data, we are giving users a pathway to immediate success, without going to the extraordinary lengths they may otherwise have to. All of this is underscored by full security, governance and transparency in how AI comes to life in Domo.”
Another detail worth a mention here is rooted in Domo’s new AI Model Management facility, which comes decked up with an ability to simplify registration and management of external models, including those hosted in OpenAI, Databricks, Amazon Bedrock, Hugging Face, and more. On top of that, you can also bank upon the stated facility to build and train models in Domo and deploy them from Domo Jupyter or AutoML. After they are successfully deployed, these models can be again leveraged from apps, workflow engine or Magic ETL, thus helping even those who aren’t exactly familiar with how to create an AI model and unlock real business impact from AI models. Complimenting the same is Domo’s decision to launch new Hugging Face models for external hosting. Our final piece of highlight is an all-new ResponsibleGPT App. Unlike other GenAI chat services, such as ChatGPT or Gemini, that add an individual’s personal information to their public LLM, Domo’s ResponsibleGPT App leverages the company’s text generation AI Service to establish an API connection with selected LLM. Furthermore, the API calls are not stored within the public LLM model, and all conversations in the ResponsibleGPT app are also stored in a Domo dataset for auditing and administrative review.
“Based on our research we assert that by 2027, three-quarters of all data processes will use AI and ML to accelerate the realization of value from the data,” said Matt Aslett, Director of Research, Analytics and Data at Ventana Research, now part of ISG.
