Finyuus
Code-first building, running, and governance of durable AI workflows
About Finyuus
Finyuus is a code-first platform for building, running, and governing durable AI workflows. It provides an indentation-based domain-specific language for composing agents, tools, guards, human approvals, structured data operations, and nested workflows; workflows run on Temporal, with support for state management, retries, cancellation, child workflows, and approval waits. The platform also provides a centralized environment for version management, observability, auditing, and operations for these workflows, and can publish them as authenticated HTTP endpoints through the Pipeline API.
Best for
Best use cases
Build durable AI workflows
Compose Agents, tools, guards, and human approvals
Manage workflow versions, observability, and auditing
Publish workflows as authenticated HTTP endpoints
Features
Indentation-based workflow language
Compose agents, tools, guards, human approvals, structured data operations, and nested workflows with an indentation-based language.
Temporal durable runtime
Manage workflow state, retries, cancellation, child workflows, and approval waits.
Pipeline API
Publish workflows as authenticated HTTP endpoints and provide validation, idempotency, rate limiting, and server-sent events.
Centralized governance environment
Centralize version management, observability, auditing, and operations for workflows.
Pros
- Code-first definition of workflows and their components
- Temporal provides state, retries, cancellation, and approval-wait management
- Supports version management, observability, auditing, and operations
- The Pipeline API provides validation, idempotency, rate limiting, and server-sent events
Cons
- Configuration requires at least one model provider key
- The global request rate for the Pipeline API is configured at 300 per minute