Artificial Intelligence is driven by Large Language Models (LLMs). These machine learning models learn and adapt algorithms and statistical models to analyze and draw inferences from patterns in data. The resulting machine learning algorithms can be leveraged via apis to get AI-powered automations in your apps.
Glide AI offers AI automation and user-triggered data transformations, such as converting audio to text, generating new text, extracting text from images, and converting text to boolean, choice, date, or number formats. These capabilities enable advanced data manipulation and feature-rich, codeless mobile app software development.
All of Glide AI’s features can be used as columns or actions, and they can be added in the data editor, the layout editor, or in the workflow editor. You can leverage AI to streamline your data and create optimizations for your team’s workflows.
Don't see Glide AI features?
Glide AI is available for paid plans only.
Features
Glide AI has nine features. They are:
Select any of the features above to view their full setup and configuration instructions. Enterprise customers also have access to an Advanced Reasoning feature. This feature is experimental and uses the most advanced models to provide more advanced responses.
A note on the accuracy of AI responses
Glide AI's models generate answers to the best of the model’s capabilities. The accuracy of these answers is not guaranteed, and we recommend reviewing the results for accuracy.
Instructions and Inputs
Instructions
Most Glide AI features have an Instructions field. Instructions are the how of Glide AI: how will the response be shaped? This is where you can provide specific directions to guide the response you get from your automation. The field is optional, and if you leave it blank the feature will use whatever Glide’s default instructions are. By adding your own instructions, you can put AI technology to work in exactly the way you need for your specific use case.
Instructions are the way to seed the AI with the personality or the context you want it to consider when responding to the message, so it will be more adaptive to your business. The default response is generic, polite, and concise. You can get more from Glide AI and tweak the flavor of the response if you specify the instructions: adjust the tone, format (bullet points, numbered list, array, comma-separated values), length, code, and more. Adding instructions can make interacting with the ai move intuitive for your users.
Inputs
Most Glide AI features have a required Input field. The input is the what for Glide AI: what input will the feature be considering as it shapes a response? What question will it answer? What text will it complete?
Depending which feature you use, the input can take many different formats. You can start from scratch with a new data table, or use your existing data in input fields to combine your dataset with the power of Glide AI. Note that the language models were built in English, so even though they have some capabilities in other languages, their accuracy may suffer.
You can also create web applications with a user interface that allows users to input text and interact with your ai automation directly. We recommend having a software testing process in place so you can discover any undesirable test results and get the best ai results.
Not sure how to get started? Check out AI templates and discover what ai-powered features can do.
Structured Responses
Often when we think about AI tools we think about what it can do with text: completing prompts with predictive text, generating emails with natural language processing, answering user questions in real-time with chatbots, and use cases that eliminate repetitive tasks. We’ve also seen lots of examples of what cloud-based, user-friendly models like openAI’s chatgpt These are all impactful automations for your business processes, but with Glide AI, you can also create high-quality structured responses from your existing data, like making your own gpt. With features like Text to Boolean, Text to Choice, Text to Date, and Text to Number, you can use AI to analyze unstructured text and return valuable data that you can repurpose throughout your apps.