Durable execution for workflows and AI agents. Self-hosted on PostgreSQL and NATS—no US cloud dependency.
@workflow(name="order-processing")
class OrderWorkflow:
async def run(self, ctx: WorkflowContext, input: OrderInput) -> OrderOutput:
# Schedule payment task (string-based task reference)
payment = await ctx.execute_task("payment-task", {
"amount": input.amount,
"currency": input.currency
})
# Schedule inventory task
inventory = await ctx.execute_task("inventory-task", {"items": input.items})
if inventory["out_of_stock"]:
await ctx.run("refund", lambda: refund_payment(payment["transaction_id"]))
raise NonRetryableError("Inventory unavailable")
return OrderOutput(order_id=input.order_id, status="completed")Everything you need to build reliable distributed workflows
Write workflows in Python, TypeScript, Kotlin, or Rust with a simple WorkflowContext API. Event sourcing ensures deterministic replay.
Resume agent loops at any point without losing state. Stream LLM tokens in real-time. Time-travel debugging lets you replay executions.
Complete audit trail of all executions. Replay workflows from any point. Understand exactly what happened and why.
Automatic retries, compensation logic, and failure handling. Focus on business logic, not infrastructure complexity.
Multi-tenancy, quota enforcement, rate limiting, and data retention. gRPC communication protocol. Deploy on your infrastructure with full control.
Flovyn handles the complexity of distributed workflows so you can focus on business logic
Process orders across payment, inventory, and fulfillment with automatic compensation on failures. If delivery fails after payment, Flovyn automatically triggers refunds and inventory restoration.
Handle failed payments with scheduled retries, notification escalation, and support team handoff. Ensures reminders aren't duplicated and maintains exact billing state for compliance.
Agent loops that search web, analyze results, decide next steps, and compile reports. If the agent crashes mid-research, Flovyn resumes exactly where it left off without losing context.
Receive webhooks from payment providers, validate signatures, enrich with database context, update downstream systems. Guarantees exactly-once processing despite webhook retries.
Process thousands of documents in parallel (OCR, ML inference, validation), aggregate results, handle partial failures by retrying specific items.
Build SaaS products with complete tenant isolation, quota enforcement, rate limiting, and team-based access control. Built-in multi-tenancy and RBAC.
Stop reinventing distributed systems primitives. Focus on business logic.
No custom storage to learn. No proprietary protocols. Your team already knows how to backup, monitor, and scale these systems.
Workflow state in Postgres. Use pg_dump, your existing monitoring, and proven scaling patterns.
High-performance task distribution. No proprietary protocol—standard NATS tooling works.
Your on-call team already knows these systems. No 3am debugging of custom storage engines.
SELECT * FROM workflows. Debug issues with tools you know. No black-box storage layer.
For organizations that need workflow orchestration without sending data to US clouds.
Designed and built in Europe by a European team. Not a US company with an EU subsidiary.
Your workflow data stays on your infrastructure. No data routing through US servers.
Built with European privacy regulations in mind. Full audit trails for compliance.
Self-hosted on any infrastructure. No mandatory AWS, Azure, or GCP dependencies.
The Flovyn server is written in Rust. No JVM, no interpreter, no garbage collector surprises.
No garbage collection pauses. Consistent low-latency workflow execution under load.
Compile-time guarantees prevent null pointer crashes and memory leaks in production.
Efficient memory footprint. Run thousands of workflows on modest hardware.
One executable + PostgreSQL. No JVM, no containers required, no runtime dependencies.
Dive deeper into how Flovyn works
Define tasks, compose workflows, and run your first worker in three steps.
Learn moreSide effects, durable state, timers, signals, and real-time streaming across all SDKs.
Learn moreBuild reliable AI agent loops that survive crashes and resume exactly where they left off.
Learn moreDeploy on your infrastructure with PostgreSQL and NATS. European sovereignty. No US cloud dependency.