Blog
Product
blog category

You don't have to worry about all the different LLM options out there

By
Mark Ku
November 8, 2024
You don't have to worry about all the different LLM options out there

The Early Stages of our LLM Router

We've added more Large Language Models (LLMs) like Claude and Gemini in addition to OpenAI because different LLMs can impact the quality of responses you receive. Right now, you're only able to trigger different LLMs, but we're building an LLM router which will select the best AI model for your specific task. Here's how it works:

  • Storytell will analyze the type and complexity of your query (e.g, reasoning, coding, creative writing)
  • Storytell will then factor in quality of response and performance
  • An appropriate LLM will automatically be selected for you

We want you to focus on what's important and doing your job well so we'll make the decision making for you. It's like having a team of AI experts working behind the scenes to give you the best answer possible.

Static Mockups of Storytell Collections

Collections will be the next thing coming to Storytell and we want to show y'all the static pages that we've made which include:

  • Improved layout with better spacing
  • Accessibility considerations, including screen reader support
  • Responsive design

We're still fine-tuning some elements like the behavior of collection items and their associated actions. Let us know what you think!

Bringing Data to Life in the UI

We've also started connecting our backend systems to the new collections UI! Watch the video bellow to check up on our progress. Here's what you can expect:

  • Three root collections where you can create your threads
  • The ability to start new SmartChats within collections

Dynamic SmartChats

Dynamic SmartChats are a feature that users from our Product Calls are excited for. They're a way to easily get updated information by having "pinned" prompts.

Chunking and Embedding

In order to make multi-document chats to work properly, we're making changes to how we process and store data like:

  • Implementing chunking strategies to break down large documents
  • Creating embeddings for similarity in the data
  • Optimizing storage and retrieval of data

Become an Alpha or Beta Tester

Become an alpha or beta tester to let us know what you'd like us to build! It's a great opportunity to shape the future of Storytell and get a sneak peek at what's coming next. Sign up here: https://web.storytell.ai/early-access

Gallery

Share this post
web.storytell.ai/blog/you-dont-have-to-worry-about-all-the-different-llm-options-out-there
linkedin
x (formerly twitter)
facebook