Inside Alibaba Cloud’s Yaochi Database: Market Outlook, AI Fusion, and Open‑Source Strategy

Re‑estimating the Size of China’s Database Market

Using publicly disclosed financial data, the interviewee derives a higher market size than Gartner’s $43 billion estimate. Alibaba Cloud reported Q1 revenue of roughly ¥250 billion. Assuming databases consume 8‑30 % of total IT spend (median 18 %, conservative 10 %), the database‑related portion of Alibaba Cloud’s revenue is ¥25‑45 billion. IDC and Gartner reports place Alibaba Cloud at 44 % of the Chinese public‑cloud market, which implies a public‑cloud database market of at least ¥200 billion. Applying the industry heuristic that public‑cloud accounts for 1/5 to 1/10 of the total (offline IDC) market yields a total Chinese database market of ¥1‑2 trillion, far exceeding the Gartner figure.

Why More Than 200 Domestic Database Vendors Can Co‑exist

Most Chinese vendors are small teams (<50 engineers). Sustainable players need deep kernel expertise (3‑5 years of database engine development) and strong technical leadership. Open‑source lowers entry barriers, but the core technology remains complex. Success depends on:

Robust engineering practices and rapid bug‑fix cycles.

Ability to grow an active user base that stresses the product.

Ecosystem support (tooling, monitoring, migration aids).

Yaochi (Alibaba Cloud Database) Brand Vision and Cloud‑Native Evolution

The “Yaochi” name (a treasure pool) reflects the goal of making data universally accessible and reducing technical friction. The product line follows a progressive cloud‑native roadmap:

Cloud‑Native 1.0 : Compute‑storage separation and elastic provisioning.

Cloud‑Native 2.0 : Fine‑grained compute‑memory separation, tighter resource control.

One‑Stop Experience : A single Yaochi endpoint can serve OLTP, OLAP, and vector workloads, automatically routing queries to the appropriate engine.

The roadmap is organized into four pillars:

Cloud‑Native 2.0 – deeper decoupling of resources.

Integration – unified access to heterogeneous workloads.

Platformization – self‑service APIs, plug‑in architecture.

Intelligence – embedding large‑model AI for automation and insight.

AI‑Driven Vector Database Capabilities

Large‑model breakthroughs have revived vector databases, moving them from niche to mainstream. Alibaba Cloud aims to embed vector capabilities in 40‑50 % of Yaochi products, enabling storage and similarity search over structured, semi‑structured, and unstructured data.

Two AI‑for‑DB initiatives are highlighted:

AI for DB – autonomous operations (self‑monitoring, self‑optimizing, self‑repairing) that reduce DBA workload.

DB for AI – integration of vector engines and large‑model services (e.g., ModelScope) to provide NL‑to‑SQL, and future NL‑to‑BI capabilities.

PolarDB Technical Innovations

PolarDB is offered in three variants:

PolarDB for MySQL (MySQL‑compatible).

PolarDB for PostgreSQL (PostgreSQL‑compatible).

PolarDB‑X (distributed “Xscale” version for massive scale and multi‑master workloads).

Key innovations include:

Three‑layer decoupling – separate compute, storage, and memory layers for flexible resource allocation and cost efficiency.

Integrated HTAP – hybrid transactional/analytical processing eliminates data movement between OLTP and OLAP.

Serverless pricing – users pay only for active compute and storage; idle resources incur zero cost.

Hardware‑software co‑design – custom compression algorithms achieve high compression ratios while maintaining low latency.

Transparent architecture transition – workloads can shift from centralized to distributed deployments without code changes.

Built‑in AI features – NL‑to‑SQL, autonomous tuning, and vector search are native capabilities.

Growth Milestones and Market Impact

PolarDB entered public beta in 2017. Highlights:

800 % growth from 2017 to 2018 – the fastest growth among Alibaba Cloud databases.

Revenue surpassed ¥1 billion in 2019.

Annual growth >50 % (2020‑2022), with >80 % in some years.

Over 6,000 cloud customers (excluding offline deployments) by 2022.

Open‑Source Strategy and Community Engagement

Alibaba contributes to major open‑source projects: AliSQL – Alibaba’s MySQL fork.

Upstream contributions to PostgreSQL and Redis (e.g., substantial patches to Redis 7.0).

The “cloud‑first, then open‑source” model means new features are first validated internally for 3‑6 months, then released to the community, ensuring a single code base for both cloud and open‑source editions.

Future Outlook for Sustained Leadership

The interviewee outlines five pillars to keep Alibaba Cloud at the forefront of database technology:

Open collaboration with academia, research institutes, and industry peers.

Competitive monitoring to stay aware of peer innovations.

Demand‑driven development – prioritize features that solve real‑world workloads.

Hardware‑centric innovation – integrate GPUs, FPGAs, and OSS object storage for performance gains.

Talent cultivation – nurture engineers with deep kernel expertise and encourage ecosystem participation.