Important Upcoming Features:

  • Hydra Chat: Fetch converasation on startup + ability to delete conversation + fix issues when stream fails due to stream gateway closing.
  • Source citations for information fetched from Web Sources, PDFs, and Internet so LLM remains truthful.
  • Build and test a comprehensive drug search, drug-drug, drug-food interaction tool for clinics with citations.
  • Enable translation on streaming responses.

​
Hydra Chat Terminal for Enterprises

This is the first iteration of our Hydra Chat Terminal that will allow for seamless interrogation of operations, access to data, and dynamic UI for displaying complex information such as graphs, tables, data sources, plethora of tools all the while remaining within this terminal.

​
40% Increase in Response Speed

Streaming: This allows responses to be streamed as AI agents think and formulate an answer, we show it as its being generated. This drastically reduces latency, where response times become a lot faster.

Improved short-term memory & data indexing strategy: There were many operations being done the moment the user asks a question, but now there are optimizations in place to help improve response times by 13s.

​
Some Controls During Response Generation

We’ve implmented a few controls to start with to allow users to cancel generation and see the answers as they are being generated.

Release 1.4Release 1.2