Build Agentic workflows, RAG pipelines, with rich AI model and tool support on one collaborative workspace. Deploy on cloud, VPC, or self-hosted, so teams move from prototype to production without rebuilding the stack.
Python agentagentic-aiagentic-frameworkagentic-workflowaiautomationclaudedeepseekgenaigptllmlow-codemcpnextjsno-codeopenaipythonskillsworkflow
from sqlalchemy import Engine
from sqlalchemy.orm import Session, sessionmaker
_session_maker: sessionmaker[Session] | None = None
def configure_session_factory(engine: Engine, expire_on_commit: bool = False):
"""Configure the global session factory"""
global _session_maker
_session_maker = sessionmaker(bind=engine, expire_on_commit=expire_on_commit)
def get_session_maker() -> sessionmaker[Session]:
if _session_maker is None:
raise RuntimeError("Session factory not configured. Call configure_session_factory() first.")
return _session_maker
def create_session() -> Session:
return get_session_maker()()
# Class wrapper for convenience
class SessionFactory:
@staticmethod
def configure(engine: Engine, expire_on_commit: bool = False):
configure_session_factory(engine, expire_on_commit)
@staticmethod
def get_session_maker() -> sessionmaker[Session]:
return get_session_maker()
@staticmethod
def create_session() -> Session:
return create_session()
session_factory = SessionFactory()