The AI ​​process orchestration engine decomposes complex business logic into ordered nodes through visual flow charts, and supports manual approval intervention, conditional routing and runtime variable tracking.

Flowra

Flowra is an AI engine focused on process orchestration. Its core value lies in modeling complex cross-system and cross-role business processes into visual flow charts and automatically advancing the execution of each node. Different from general workflow tools (such as n8n, Make), Flowra emphasizes the "Human-in-the-Loop" approval mechanism and runtime variable tracking capabilities, which is more suitable for enterprise-level scenarios with high compliance requirements. It serves as "lightweight BPM" in small and medium-sized enterprises, filling the market gap between traditional heavy-duty BPM systems and simple automation tools.

Flowra is open sourced by ModelScope and the Wuli WULI team. It is built based on the FlowBench core engine and provides a graph execution engine and node package development tool chain. The core delivery form is an open source engine + visual flowchart editor, which serves small and medium-sized enterprises by lowering the threshold for process automation and approval flow.

Core parameters and statistics

Parameter item	Value
Number of node types	25+ (AI tasks, manual approval, HTTP requests, conditional routing, timers, etc.)
Flow chart nesting level	Up to 5 levels (supports sub-process expansion)
Runtime variable tracking	Support (view the input/output variables of each node in real time)
Manual approval timeout setting	Supported (configurable from 1 hour to 7 days)
Approval Channels	Web Panel / Email / Feishu / Business WeChat
Execution history audit log	Supported (can be exported to CSV/PDF)
Maximum number of concurrent processes	20 (Free) / 100 (Pro) / 500 (Enterprise)
Open source license	Open source (subject to the official repository)
Project ownership	ModelScope (Alibaba) × WULI

Parameter interpretation: 25+ Although the node types are not as rich as the general automation platform, the design focuses more on business scenarios that "require manual judgment" - approval nodes, timeout transfer, and conditional routing are the core requirements of enterprise processes. The stepped setting of the concurrency limit means that the Pro version (100 concurrent items) can cover the daily approval volume of most growing enterprises.

User and market recognition

Flowra is currently in the early stage of public beta, with about 3,500 registered users, mainly from three industries: financial technology, enterprise services and cross-border e-commerce. Among them, the approval flow templates (purchase approval, contract approval, expense reimbursement) have the highest number of downloads, accounting for 42% of the total template usage. In the user NPS survey (sample size 200), "the flow chart is intuitive and easy to understand" received 4.5/5, and "approval channel coverage" received 4.2/5.

From the perspective of market positioning, Flowra directly targets the approval modules of traditional BPM vendors (Lanling, Zhiyuan, Panwei), but its open source + SaaS model significantly lowers the procurement and implementation threshold. Typical feedback from a user is: "The traditional OA approval process needs to be scheduled by the IT department from request to launch, with an average wait of 2-4 weeks; with Flowra, I can complete it myself in 30 minutes." Compared with Zapier or Make, Flowra's advantage lies in its deep adaptation to domestic approval scenarios - it supports IM approval channels such as Feishu and Qiwei, as well as audit log formats that comply with domestic financial regulations.

Cost advantage

Cost Dimension	Description
Free version	¥0/month, 3 active processes, up to 15 nodes per process, basic node type
Pro version	¥99/month (annual payment ¥79), 20 active processes, full node type, approval channel integration, audit log
Enterprise Edition	¥399/month, unlimited process number, sub-process reuse, SSO integration, privatized deployment consulting, exclusive SLA

Comparison with competing products:

Comparison dimensions	Flowra (Pro)	Traditional BPM platform	Self-developed approval system
Deployment method	SaaS, zero deployment	Local deployment, operation and maintenance required	Development team required
Time to build a procurement approval flow	About 30 minutes	2-5 days	2-4 weeks
Monthly fee	¥99/month	¥3,000+/month (License)	R&D labor cost
Modification process flexibility	Drag and drop modifications to take effect in real time	Change application is required	Development needs to be scheduled
Audit logs and compliance	Built-in	Perfect	Need to build by yourself

Hidden cost reminder: In the SaaS model, data is stored in the cloud. Industries such as finance and government affairs that have clear requirements for data residency need to confirm whether the enterprise version privatized deployment plan meets compliance requirements. The efficiency of the approval flow depends on the response speed of the approver - no matter how fast the process engine is, the failure of the approver to process in a timely manner will still extend the overall cycle.

Main functions

• Visual flowchart editor: Drag-and-drop canvas, supporting swim lane diagrams (divided by role/department), node connections, color labels and annotations. Based on virtual node rendering technology, the flow chart of 200+ nodes maintains 60fps smoothness.
• Manual Approval Node: Supports single-person approval, countersigning (all agree to pass), or signing (any one agrees to pass), and sequential approval (approval in sequence). Supports automatic timeout transfer and reminder reminders.
• Conditional Routing and Parallel Branching: Dynamically select subsequent paths based on form field values ​​or node output results. Parallel branches support advancing multiple independent tasks simultaneously.
• Subflow encapsulation: Encapsulate fixed process segments into reusable subflow graphs, with independent variable scopes, and exchange data through input/output mapping.
• Runtime Variable Panel: View the input/output data, time consumption and status of each node in real time during process execution.
• Form Engine: Drag and drop to build approval forms (text box, radio selection, multi-selection, date, attachment, rich text), providing 20+ preset form templates.
• Statistical dashboard: Generate execution statistical reports based on process type, processor, time consumption, timeout rate and other dimensions.

Model and version evolution

Version	Release Date	Major Changes
v0.9 (Preview)	2026-06-10	Supports linear flow charts, basic node types (start/end/task), excluding approval and routing
v1.0 (Public Beta)	2026-07-14	Completely introduces approval nodes, conditional routing, sub-processes, form engines, and statistical dashboards

From preview to public beta, Flowra has completed the leap from "flowchart drawing tool" to "executable process engine". The preview version only supports linear processes, and the biggest functional increment of the public beta version is the combination of approval nodes and form engines - making Flowra practical in real enterprise scenarios.

Technical advantages

• Process State Machine Kernel: Use the state machine model to drive process execution, and each node corresponds to a clear status (to be executed/executing/completed/failed/timed out/skipped). The state transition rules are solidified in the process definition stage, and the process status and location can be clearly answered at any time.
• Variable scope isolation: Subprocesses have independent variable scopes and exchange data with the main process through explicit input/output mapping to avoid variable name conflicts in large processes.
• Automatic arbitration after approval timeout: When the approver times out and fails to process, arbitration will be automatically performed according to preset rules (transferred to superior/automatically skipped/automatically rejected), and the arbitration action will be fully recorded in the audit log.
• Incremental Rendering Engine: The flowchart editor uses virtual node technology, with 200+ nodes maintaining 60fps dragging smoothness.

How to use

How to use	Entrance	Instructions
Web Flowchart Editor	Official website login → Process Design	Create and edit flowcharts
Initiate process	Workbench → Initiate process	Select the published process template, fill in the form and submit
Approval task processing	To-do list / email / IM notification	Approval processing through various channels
Operation Monitoring	Monitoring Center → Running Instance	View the real-time execution status of each process

Typical usage process: Design flow chart → Configure node parameters → Bind form template → Release process → User-initiated process → System automatic flow → Approval processing → Continue flow/end → Archive audit.

Product Pricing

Package	Price	Core Benefits
Free version	¥0/month	3 active processes, up to 15 nodes per process, basic node type
Pro version	¥99/month (annual payment ¥79)	20 active processes, full node types, approval channels (Feishu/Qiwei), audit logs
Enterprise Edition	¥399/month	Unlimited number of processes, sub-process reuse, SSO integration, privatized deployment consulting, exclusive SLA

New users can try the Pro version for free for 14 days during the open beta. The free version for teams under 50 people can cover 3-5 core approval flows.

Application scenarios

• Purchase Approval Process: Employee submits purchase application → Department supervisor approval → Financial budget verification → Procurement execution → Arrival confirmation → Financial payment. The entire process is traceable and the cycle time is shortened from 3-5 days to 1 day. Verification method: Compare the average approval cycle before and after deployment.
• Contract countersigning and archiving: Legal drafting contract → Relevant departments will countersign in turn (leaving traces of approval opinions) → Legal final review → Electronic signature → Automatic filing. Verification method: Randomly check the completeness and timely filing of countersigning records.
• Employee Onboarding Process: HR initiates onboarding → parallel branches notify IT, administration, and departments at the same time → summarize after completion of each branch. Verification method: Comparison of the timeliness of completion of new employee onboarding preparations.
• Customer refund processing (human-machine collaboration): Customer service inputs refund application → AI automatically verifies refund conditions → Automatic approval processing of qualified items → Manual intervention for abnormal orders. Verification method: accuracy rate of automatic approval and proportion of manual intervention.

Applicable people

Crowd	Adaptation value	Preconditions
Enterprise operation/administration	Build standardized approval flow and business process	Just understand the business rules
Project managers	Visualize project stages and key nodes through flow charts	No technical threshold
Production and research team	Cross-system coordination process automation	Understand the concepts of API and Webhook
Compliance and risk control personnel	Use audit logs to meet compliance checks	Requires Enterprise Edition features

Unsuitable scenarios: Real-time transaction processes that require extremely high throughput (such as payment processing systems) - manual approval mode adds inevitable delays; large enterprises that have deeply used traditional OA systems and have more than 50 processes - the current coverage may not be enough.

Summary and Outlook

Flowra's core competitiveness lies in the organic combination of "manual approval", the most common coordination action in enterprise management, with automated flow charts. It is more in line with the compliance and internal control needs of Chinese companies than purely automated tools - complete end-to-end automation is unrealistic in many scenarios (compliance requirements require manual approval), and Flowra's "human-machine collaboration" model is more down-to-earth at this stage.

Current limitations: (1) The SaaS model is limited in very large enterprises with strong demand for privatized deployment; (2) The integration ecosystem (especially mainstream domestic ERP/OA systems such as UFIDA and Kingdee) still needs to be expanded; (3) The commercial support system for open source projects is still under construction. Procurement/Adoption Suggestions: Before purchasing, you need to confirm whether the core process supports exporting as offline backup (to prevent supplier risks), as well as the specific implementation plan and technical support SLA for the privatized deployment of the enterprise version. Directions for subsequent versions are expected to include approval SLA statistical analysis, mobile approval experience optimization, and richer third-party connectors.

Related tools: CrewAI, langchain

Architecture design and technology selection

As an open source project, Flowra's architecture design, community health, and operation and maintenance maturity are core dimensions that need to be comprehensively considered when selecting technology. The following is a systematic framework for assessing the production readiness of open source projects.

Architecture and Modular Design The architectural design of the project directly determines the flexibility of secondary development and integration. Projects that adopt microservices, plug-in or event-driven architecture usually have better scalability and functional isolation, making it easier for the team to expand and customize specific modules on demand; the monolithic architecture is simple to deploy, intuitive to operate and maintain, and is suitable for small-scale use and rapid verification. However, as functions increase, they may face problems such as increased maintenance complexity and accumulation of technical debt. It is recommended to read the project's architecture documents and developer guides before selecting, and evaluate the adaptability of the architecture design to the team's existing technology stack, as well as the scalability of the architecture as business grows in the future.

Community health and long-term maintenance The community health of an open source project is a key indicator of whether the project can be maintained and developed over the long term. It is recommended to comprehensively evaluate the following dimensions: the growth trend and absolute value of GitHub Stars (reflecting community attention and user base), the number and composition of contributors (the ratio of core maintainers to temporary contributors, ideally there are at least 3 active core maintainers), the median issue response time (ideally within 24 hours, reflecting the response efficiency of the maintenance team), PR merge rate and merge delay (reflecting the standardization and efficiency of project governance), and the time of the latest major Release (more than 6 Months without updates should be taken as a sign that project maintenance is stalled). An active community means faster bug fixes, more frequent feature updates, a richer third-party integration ecosystem, and it’s easier to get help from the community when you encounter problems.

Deployment, operation and maintenance and production readiness Production environment deployment needs to focus on evaluating the following aspects: the completeness of the Docker image and version labeling strategy (whether multi-architecture mirroring is provided), the availability and document quality of one-click deployment scripts (docker-compose, Helm Chart, Terraform, etc.), the number and management complexity of runtime dependent components (the more dependencies, the complexity of operation and maintenance increases exponentially), the integration support of monitoring and logging infrastructure (Prometheus indicator exposure, Grafana dashboard, structured log output), and complete documentation of backup, recovery, and high-availability solutions. It is strongly recommended to go through the entire deployment process in the test environment, strictly follow the documentation from scratch, verify the accuracy of each step and the compatibility of the environment, and put it into production after all functions have been verified.

Version Info

• Public beta version ：The public beta version supports visual flow chart modeling, conditional branch routing, manual approval nodes and runtime variable panels. （2026-07-14）
• preview version ：Technology preview version, basic DAG execution engine, supports image, video, audio and other multimedia data processing. （2025-12-01）