​
How it works

  • Prompt CMS - store & version prompts; useful for deploying prompt changes without requiring a code change.
  • Prompt Playground - test new prompts variants and run evals against them using datasets.
  • Prompt Optimizer - automatically generate the best prompt for your task.
  • Evaluations - test changes of your prompt, RAG & AI agents before deploying them.
  • Monitoring - track token usage, latency & errors of your AI application in real-time.

For more details, check out Concepts.

​
Setting up

Create a Hamming.ai account or sign in. Navigate to the Settings page and Create new secret key, optionally naming the key. Make sure to save this somewhere safe.

Prompt CMS

Store, version, and deploy prompts without code change.

Prompt Playground

Test new prompts variants and run evals against them using datasets.

Prompt Optimizer

Automatically generate the best prompt for your task.

Evaluations

Get started with running evaluations.

Monitoring

Setup real-time monitoring.

Concepts