Subscribe for updates
Back to Changelog
October 2024 Product Update

October 2024 Product Update

On-premise, Global Analytics, Linear Flow, Prompt Editor, Cerebras, and more.

Paul Omenaca

Paul Omenaca

@houmland

October has been packed with updates. We’ve introduced new features, boosted system performance, and welcomed new team members.

Let’s dive into the product changes.

Stack AI On-Premise


On-premise deployment

Stack AI is now available on-premise. Host Stack AI in your own cloud (AWS, GCP, Azure, Kubernetes, etc.).

Keep your data within your own infrastructure, ensuring maximum control and compliance.

If you believe on-premise is the right choice for your needs, learn more here.

Cerebras


New AI model - Cerebras available in Stack AI

Cerebras is a new LLM provider available in Stack AI that offers high-performance AI inference capabilities. It aims to provide ultra-fast inference to enable more responsive AI applications.

It delivers up to 2,100 tokens per second for Llama 3.1 70B model inference, significantly faster than competitors:

  • 16x faster than the fastest GPU solution
  • 8x faster than GPUs running Llama3.1-3B, a model 23x smaller

If you are curious about this ground-breaking provider, read more here.

Global analytics


Global analytics

Analyze what’s going on in your organization with the global analytics tab in your dashboard.

Each user will see analytics of the projects they have access to only. An admin user will be the only only seeing all projects.


Global analytics tab zoom in

Filter projects by name or by metric, and access the logs of a specific project directly from this tab.

Linear flow


Linear flow

Introducing a new way to work with information across your projects - Linear Flow!

Now, nodes connected together can access previous node data automatically.

For example, there’s no need to re-link your input node to the LLM if it's already connected to a knowledge base.


Linear flow example

More features ⚒️

Check out other features we’ve released this month:

  • Microsoft Synapse Integration – Seamlessly connect with Synapse
  • Llama 3.2 Models – Access Meta and Together AI models
  • Rename Nodes – Now available for input nodes
  • Add Comments – Annotate nodes for easier collaboration
  • New Prompt Editor – Just type ‘/’ to view available information from previous nodes—no more need for brackets

New Content

Check out our latest video and blog post releases: