Monitoring

Literal AI has a Dashboard on which you can monitor usage of your application.

​

Metrics

Various charts inform you about:

• Performance: measure latency and token throughput per second.
• Quality: LLM Evaluation: view Human feedback scores, eval experiment scores and product metrics
• Cost: see the token usage per model in a time window.
• Volume: follow the number of threads developing over time, the number of (active) users, the amount of feedback and the token usage.

These metrics can be displayed over a specified time window.

Literal AI Dashboard