Ziqi MainData Science at MicrosoftMany facets of un-truth: LLM hallucination 101With the increased adoption of LLMs across industries, hallucination has become one of the hottest related buzzwords — after all, when one…Sep 5, 2023Sep 5, 2023
Ziqi MainData Science at MicrosoftReasoning about change: Lessons learned from building a near real-time system for Azure pricingTraditionally, when we think about data, we think about states — static measurements or records of phenomena that we care about, in a…Apr 25, 2023Apr 25, 2023
Ziqi MainData Science at MicrosoftSpeeding up “Reverse ETL”Talking about the speed of data can be a little tricky. Nowadays data serves so many different use cases, each with its own definition of…Mar 7, 20231Mar 7, 20231
Ziqi MainData Science at MicrosoftWhat do we talk about when we talk about ML robustness?After finishing a great one-time ML analysis, it is natural to expect it to lead to a reliable, permanent solution. However, this…Dec 21, 20212Dec 21, 20212
Ziqi MainData Science at MicrosoftNavigating the (long and winding) road from innovation to productionFinding an innovative solution that enables you to solve a real problem is fascinating — perhaps one of the biggest reasons for people to…Aug 24, 2021Aug 24, 2021
Ziqi MainData Science at MicrosoftData science perspectives on averages and related areasMore knowledge can help you use these tools more effectivelyJun 8, 2021Jun 8, 2021