Settings

Configure system settings and credentials

General Settings

LLM API Credentials

Configure API keys for language model providers

Note: API keys are stored for this platform installation and are never exposed in API responses.

Database Connections

Configure database connection strings

PostgreSQL

Unknown

Neo4j

Unknown

Redis

Unknown

Memory Service (Graphiti)

Configure long-term memory integration with Neo4j graph database

Enable Memory Service

Activate Graphiti for long-term memory and knowledge graphs

Unique identifier for grouping related memory entries

Rate Limiting

Control concurrency and throughput limits

4
120

Number of parallel trial executions. Higher values increase throughput but consume more resources.

60
10300

Maximum number of jobs processed per minute by BullMQ. Controls overall experiment throughput.

100
101000

Maximum API requests to LLM providers per minute. Stay within provider rate limits to avoid throttling.

Important:

Changes to rate limits will take effect after the next worker restart. Adjust carefully to balance throughput with resource constraints and provider limits.