The following page may contain information related to upcoming products, features and functionality. It is important to note that the information presented is for informational purposes only, so please do not rely on the information for purchasing or planning purposes. Just like with all projects, the items mentioned on the page are subject to change or delay, and the development, release, and timing of any products, features or functionality remain at the sole discretion of GitLab Inc.
The AI Framework group is a team focused on increasing the pace of adoption and innovation with advanced AI/ML technologies while ensuring quality. The team is developing tools that other engineers and product people can utilize to enrich their product offering.
The introduction of new capabilities and technologies in the AI/ML space, specifically with the widespread use of LLM’s created a tectonic shift in customer expectations. However, these new capabilities do not change an old truth: teams still need to obsess over their customers' needs and have a deep understanding of their problems, so they can create solutions best suited to their use cases.
To enable teams to focus exactly on these things, our aim is to remove the technical heavy lifting from the equation, so as an organization we can focus on results for customers.
AI is moving at an ever-increasing pace, and the responsibility of the AI Framework group is to ensure that it will not only keep up but will lead the way, with novel and innovative approaches to create efficiencies across the entire DevSecOps cycle. Our implementation of the Abstraction layer is designed to enable a plug-and-play infrastructure that brings AI as a service to the GitLab platform.
We aim to:
To ensure our AI features provide with a high standard of performance and security, we will iterate on the Abstraction layer. AI Framework will execute on the best ways to deploy the AI Gateway to set the market expectations around privacy and security when it comes to AI powered capabilities.
AI agents are the future of workflow automation and scale. We aim to provide the tools to create AI agents that can be used across the DevSecOps cycle. These agents will have access to relevant data to create a rich context, as well as the ability to access and use tools.
RAG and embeddings are important parts of any infrastructure that supports AI powered capabilities. They serve as powerful tools to enhance feature quality and ensure user satisfaction. By incorporating these elements into the AI usage workflow, we can ensure that our AI offerings scale effectively with our customers and teams. In the age of the LLM, as compute is becoming more accessible, it is crucial to make sure that we compliment these powerful algorithms with the right context. Our goal is to enable AI powered experience that solves for use-cases and real world problems that our customers, community and teams are facing. With that in mind, it is extremely important to ensure that our infrastructure is built for that purpose.
To ensure the responsible use and development of AI, we want to make sure that both in terms of proper usage and in terms of costs, we have a robust and reliable ways to track and measure the use of AI. Our focus here is to create it with the highest standard of user privacy. We do not track specific user activity nor we use any data to train and improve the models.
This group consists of the following category:
Last Reviewed: 2024-01-30
Last Updated: 2024-01-30