One of the original pandas authors is now building the data layer for images, video, and embeddings that tabular tools were never designed for.
Chang She is co-founder and chief executive of LanceDB, a San Francisco data infrastructure company behind the open-source Lance columnar format and a multimodal lakehouse for AI data.
Joining Wes McKinney on pandas
She began his career as a financial quant with stints at AQR and Barclays, according to his Y Combinator profile. In 2012 he joined Wes McKinney to work on pandas full time and became one of the library's original co-authors and its second major contributor.
Leading Engineering at Tubi
He co-founded DataPad, a visual analytics startup where he was chief technology officer, and the team joined Cloudera in 2014, as AI Engineer notes. He later became vice president of engineering at Tubi, designing its machine learning stack and experiment platform; Fox acquired Tubi in 2020, according to Forkable.
Starting Over on Multimodal Data
He started LanceDB with Lei Xu, a former Cloudera colleague and HDFS contributor, because conventional tabular systems handled images, video, and embeddings poorly. The company launched as a vector database, then widened into a lakehouse that combines search with training data curation. In June 2025 he announced a $30 million Series A led by Theory Ventures.
Personal Life
She lives and works in the San Francisco area. He has described his background as open source data tooling, and LanceDB keeps the Lance format open source as the base of its commercial product.
