BLUF: This milestone focuses on tackling tech debt to ensure we provide a solid, spec-compliant, and well-documented API to serve as a runtime platform for generative AI applications.

Core User Stories:

As an AI application developer who needs to bring my app to an air-gapped environment
I want LeapfrogAI to provide an API that mirrors the OpenAI API
So that the application I developed on the OpenAI API will work with LeapfrogAI.

As an AI application developer building a mission-specific application to run on LeapfrogAI
I want clear documentation for building, deploying, and using the LeapfrogAI API
So that I can easily run a local instance of LeapfrogAI to build my application.

MUST HAVE (Non-negotiable product needs that are mandatory for the team)

• All currently implemented OpenAI endpoints respond in a similar manner to OpenAI's managed service
  • Implement conformance testing to validate our API functionality against OpenAI services
• Revise/update GitHub READMEs and LeapfrogAI doc-site

SHOULD HAVE (Important initiatives that are not vital, but add significant value)

• Decrease build and e2e test times to shorten the build-test feedback loop
• Implement a backup approach/system for the database
• Publish registry1 and upstream flavors of LeapfrogAI

COULD HAVE (Nice to have initiatives that will have a small impact if left out)

• Other improvements to the general robustness of our database?
• Publish unicorn flavor of LeapfrogAI
• Complete MVP run of evals
  • Update our default model(s) based on results
  • Publish results and our model recommendations in the documentation

WILL NOT HAVE (Initiatives that are not a priority for this specific time frame)

• TBD