Configuration

Environment settings for the Quint API server

Configuration

All settings are managed through Pydantic Settings loaded from environment variables (.env file or system env).

Settings Reference

Database

Variable	Default	Description
DATABASE_URL	postgresql+asyncpg://quint:quint_local_dev@localhost:5432/quint	Async PostgreSQL connection string

Redis

Variable	Default	Description
REDIS_URL	redis://localhost:6379/0	Redis connection string
REDIS_EVENT_TTL	86400	Event cache TTL (seconds)
REDIS_POLICY_TTL	3600	Policy cache TTL (seconds)
REDIS_MAX_EVENTS_PER_CUSTOMER	10000	Max cached events per customer

API Server

Variable	Default	Description
API_HOST	0.0.0.0	Server bind address
API_PORT	8000	Server port
API_WORKERS	4	Uvicorn worker count
CORS_ORIGINS	["http://localhost:3000"]	Allowed CORS origins

Rate Limits

Variable	Default	Description
RATE_LIMIT_STARTER	10000	Events/day for starter tier
RATE_LIMIT_PRO	100000	Events/day for pro tier
RATE_LIMIT_ENTERPRISE	1000000	Events/day for enterprise tier

GraphReasoner

Variable	Default	Description
GRAPH_REASONER_ENABLED	True	Enable primary scoring engine
GNN_ENABLED	False	Enable GNN structural scoring
GNN_MODEL_PATH	""	Path to trained GNN model
GNN_WEIGHT	0.0	GNN weight in ensemble
FORWARD_CHAIN_WEIGHT	1.0	Forward-chaining weight

Memgraph

Variable	Default	Description
MEMGRAPH_URL	bolt://localhost:7687	Memgraph Bolt URL
MEMGRAPH_USERNAME	""	Memgraph username
MEMGRAPH_PASSWORD	""	Memgraph password
MEMGRAPH_ENABLED	False	Enable Memgraph
MEMGRAPH_ENRICH_ENABLED	False	Enable enrichment queries
MEMGRAPH_EVENT_STORE_ENABLED	False	Write events to Memgraph
MEMGRAPH_QUERY_TIMEOUT_MS	30	Query timeout (ms)

LLM (Gemini)

Variable	Default	Description
GEMINI_API_KEY	""	Google Gemini API key
GEMINI_MODEL	gemini-2.5-flash	Model name
GEMINI_TIMEOUT	30.0	Request timeout (seconds)

Risk Engine (gRPC)

Variable	Default	Description
RISK_ENGINE_ADDRESS	localhost:50051	gRPC server address
RISK_ENGINE_TIMEOUT	120.0	gRPC timeout (seconds)
RISK_ENGINE_ENABLED	False	Enable GPU risk engine

Modal (GPU)

Variable	Default	Description
MODAL_TOKEN_ID	""	Modal authentication token ID
MODAL_TOKEN_SECRET	""	Modal authentication secret
WEIGHTS_VOLUME_NAME	customer-lora-weights	Modal volume for model weights

Orchestrator

Variable	Default	Description
BATCH_INTERVAL_MINUTES	5	Batch processing interval
MAX_EVENTS_PER_CHUNK	256	Max events per processing chunk
MAX_PARALLEL_CHUNKS	10	Max parallel chunk processing
MAX_RETRY_COUNT	3	Max retries before dead letter

Caching

Variable	Default	Description
POSTGRES_CACHE_TTL_DAYS	7	PostgreSQL L2 cache TTL

Archival

Variable	Default	Description
ARCHIVE_AFTER_DAYS	90	Days before archival to S3
ARCHIVE_STORAGE_BUCKET	quint-archive	S3 bucket name
ARCHIVE_STORAGE_ENDPOINT	""	S3-compatible endpoint

Logging

Variable	Default	Description
LOG_LEVEL	info	Logging level
ENVIRONMENT	local	Environment name (local/staging/production)

Local Development

Create a .env file in the infra/ directory:

DATABASE_URL=postgresql+asyncpg://quint:quint_local_dev@localhost:5432/quint
REDIS_URL=redis://localhost:6379/0
GEMINI_API_KEY=your-key-here
GRAPH_REASONER_ENABLED=true
LOG_LEVEL=debug
ENVIRONMENT=local

Start backing services and run:

docker compose up -d postgres redis
python scripts/migrate.py
python scripts/seed.py
uvicorn src.api.main:create_app --factory --reload --port 8000