Atlas Platform for API Governance and Knowledge Management at iQIYI
iQIYI’s Atlas platform consolidates API governance and knowledge management by providing a unified portal with OpenAPI 3.0 specs, one‑click debugging, link tracing, knowledge graphs, and automatic code‑as‑documentation, dramatically cutting development and testing time, reducing maintenance costs, and streamlining collaboration across its 100‑million‑member microservice ecosystem.
With iQIYI membership surpassing 100 million, the membership service evolved from a small, fast single‑business system to a domain‑segmented microservice architecture to support rapid business growth and high‑traffic calls. However, the proliferation of microservices increased inter‑system interactions, raising collaboration costs during development.
Key pain points identified include:
Heavy historical baggage and inconsistent API management tools (WIKI, YAPI, Swagger).
High maintenance cost as code and documentation must be updated simultaneously.
Expensive integration and testing due to environment and link issues, causing delays for complex requirements.
The membership team therefore needed a platform to centralize API control and improve development and testing efficiency.
Solution Research
Existing platforms could not meet the membership team’s needs. Atlas, an internally developed platform, already provides basic API management capabilities and can be extended to include features such as code‑as‑documentation, one‑click link troubleshooting, unified API specifications, and automatic generation of signatures and entities. This not only solves the membership team’s pain points but also empowers other departments, avoiding duplicated effort.
Atlas is a business‑knowledge management platform built on API management. It offers a unified API governance portal, OpenAPI 3.0 specifications, API debugging and mock, link tracing, and relationship graph capabilities, effectively addressing documentation, sharing, and collaboration challenges.
Project Goals
Simplify Development : Atlas streamlines front‑back end data interaction, enabling developers to use mock data for integration without waiting for backend progress.
Intelligent Documentation : Documentation stays in sync with code changes, achieving “code is documentation, documentation is code”.
Efficient Collaboration : Features such as one‑click debugging, link tracing, automatic code generation, and subscription‑based change notifications enhance coordination between front‑end and back‑end teams.
Solution Details
01 One‑Click Verification
By integrating with the full‑link monitoring system, Atlas provides one‑click environment and call‑chain issue location during development and testing, dramatically reducing troubleshooting time. The feature currently covers asynchronous tasks, RocketMQ, and API issues.
Architecture diagram:
02 Knowledge Graph
To visualize complex service dependencies, Atlas automatically extracts interface‑level and field‑level dependencies from the full‑link monitoring system and displays them as graphs using G6.
03 Query Mode
One‑Click Search : API docs, database schemas, error codes, and UML diagrams are searchable from the homepage, standardizing information retrieval.
Historical Trace & Version Comparison : Compare any two branches or versions of an API to understand changes and context.
Subscription & Notification : Users can subscribe to APIs and receive real‑time change alerts.
04 Enhanced Debugging
Atlas extends signature handling to support template‑based selection and custom code snippets for dynamic signature calculation, simplifying debugging with tools like PostMan.
Signature algorithm templates with one‑click selection.
Custom signature algorithms via pasted code snippets, dynamically parsed and executed.
System Value
By Q3, the platform had onboarded 75 membership‑related requirements, delivering significant efficiency gains:
Development Efficiency : Each requirement saved ~2 hours, eliminating environment/link issues and reducing API maintenance costs.
Testing Efficiency : Link‑issue troubleshooting time reduced by 93% (15 min → 1 min) and environment issue resolution time cut by 50% (4 h → 2 h).
Future Plans
Refine Debugging Features : Elevate debugging to the homepage, decoupling it from API data and integrating one‑click verification.
Automated Code Generation : Use Atlas‑provided standard JSON schemas to auto‑generate front‑end code.
UML Diagram Editing : Offer online creation and editing of various UML diagrams to further boost collaboration.
iQIYI Technical Product Team
The technical product team of iQIYI
How this landed with the community
Was this worth your time?
0 Comments
Thoughtful readers leave field notes, pushback, and hard-won operational detail here.