Observe your AI apps and cloud infra they run on with Okahu to understand how to make them work better - more reliable, performant and cost-effective.
Okahu AI Observability as a Service is now in preview with BetaOne features!
Okahu BetaOne includes:
Discovery of python AI apps that use LLMs from OpenAI or Nvidia Triton to understand what components the app uses and how they work together.
Observations of execution of app and utilization metrics from LLM providers to understand what impacts performance.
Try Okahu with your OpenAI
Explore Okahu capabilities by observing a chatbot that uses OpenAI and Langchain running in the cloud. You'll need:
An Okahu AI Observability Cloud tenant
An Okahu tenant
A Postman account to explore API examples
A sample AI app hosted in the cloud
A Github account to run the app in Github Codespaces
An OpenAI account
Alternatively, you can use your favorite REST API tool with Okahu API docs and run the sample app on your laptop or a cloud VM with internet access.
Run a sample AI app in Github Codespaces
Check out this Github repo for a sample chatbot that uses OpenAI and Langchain that you can run right from your Github Codespaces. Configure this app to use your OpenAI account and send observations to your Okahu tenant using instructions provided with the repo.
Try Okahu AI Observability with Postman API workspace
Look up your Okahu Tenant ID and API Keys to use Okahu API examples. These would be sent to the email you provided to request an Okahu tenant from [email protected].
To setup your Postman workspace
Fork the Okahu BetaOne collection into your Postman workspace.
Configure Authorization at the collection level to use Okahu API Key.
Rest of the example requests inherit auth from the parent.
Run the get tenant and compare with tenant_id in response to validate.
Now you can use the examples in the collection to explore the demo.
Check out the Okahu API docs for reference.
Get help
Contact us at [email protected] if you have any questions or need help.