Case Study: SHAREit Group streamlines AI workflows with PingCAP's TiDB and TiKV

SHAREit cuts feature engineering time to 25% with PingCAP

SHAREit Group, a global content platform with nearly 2.4 billion installs, faced significant inefficiencies in its AI workflow. Algorithm engineers were spending 50% of their time on data preparation tasks like feature engineering and sampling due to incomplete data platforms and a lack of standards, which slowed down their entire model development and deployment process.

To address this, SHAREit implemented PingCAP's TiDB and TiKV to power its new AI platform and Feature Store. Using TiKV for high-performance online storage and planning to use TiDB for its SQL interface in online learning, PingCAP's products helped streamline data preparation. This solution reduced the time algorithm engineers spent on feature engineering and sampling from 50% to 25% or less, greatly improving efficiency and engineer satisfaction.

SHAREit Group