Ming-Chuan Wu's Email Addresses & Phone Numbers

📖 Summary

Currently an Engineer Manager at MSFT Big Data team -- SCOPE/Cosmos, leading the Query Optimizer team. SCOPE is a cloud-scale distributed computation platform built on top of Cosmos, a distributed file system design for high availability and scalability. The SCOPE/Cosmos platform runs as a service over tens of thousands of commodity servers, powering Bing and other Microsoft product teams. The SCOPE system combines benefits from both parallel databases and MapReduce execution engine. The SCOPE language is a declarative scripting language, resembling SQL, with integrated .NET extensions. It provides a single machine programming abstraction so that all our Bing developers can focus on their application logic rather than dealing with low-level details of distributed computation. At the same time, it allows us to perform transparent and systematic query optimization. My main charter is to design and implement the query optimizer for SCOPE that will transform the SQL-like scripts into efficient data flow execution plans that run on a MapReduce execution environment. Specialties: - federated query processing for heterogeneous cloud-scale data sources - dynamic query optimization - cloud-scale distributed computation platform - query optimization for cloud-scale MapReduce environments - indexing strategies at MapReduce environment - Large scale testing infrastructure with automatic scaling - workload analysis and physical data design for big data - query optimization for relational databases - parallel database - bitmap indexingPrinciple Engineer @ Working on an incubation project - web scale graph database engine in the cloud. From July 2015 to Present (4 months) Greater Seattle AreaPrincipal Software Engineer Manager (MSFT Big Data Infrastructure) @ Lead the Query Optimizer team, and contribute to the following projects: - Query Processing for federations of large scale heterogeneous data clusters - Dynamic Query Optimization - Query Optimization in large scale clusters - Query Processing and Optimization in Parallel Databases and Database Clusters From January 2014 to Present (1 year 10 months) Greater Seattle AreaPrincipal SDE (Bing Infrastructure team) @ Core Search (Bing), Infrastructure Team - Indexing strategies and workload analysis for big data - advanced join strategies for big data analytic workloads - Automatic release quality validation (leveraging SCOPE's distributed computation platform to validate SCOPE's release candidate on previous workload, c.f., http://research.microsoft.com/en-us/um/people/jrzhou/pub/Playback-DBTest12.pdf) - Large scale data processing and optimization; - Relational query optimization in the cloud; - SQL-like query processing and optimization in Cosmos/SCOPE (Structured Computations Optimized for Parallel Execution) From October 2008 to Present (7 years 1 month) Senior SDE (SQL Server, Storage Engine Team) @ - Development of a variety of features in the area of Parallel Query Processing, including parallel plan generation, parallel query execution, automatic data partitioning to maximize the parallelism, etc. in a shared-nothing database cluster (patent pending) - Design of distributed transaction deadlock monitor in SN environment (patent pending) - Design of a communication subsystem for distributed operations for future releases of SQL Server From August 2005 to October 2008 (3 years 3 months) Software Design Engineer (SQL Server, Query Optimizer Team) @ Development of a variety of features in the area of Query Optimization, including statistics on view, quality matrix of the cardinality estimate, quality matrix for statistics selection, improvement of cardinality estimates for complex join predicates (patented), etc. From October 2001 to August 2005 (3 years 11 months) Researcher @ Design the bitSQL database engine which exploits bitmap indexing for query processing in Data Warhouses, including an automatic query rewrite to transform operators in relational algebra into operators in Boolean algebra. This optimization technique turns costly relational operators into a sequence of efficient bit-wise operations, and most of the query processing can be done on the bitmap indexes without accessing the base tables to reduce response time. Design the extension of a Real-Time Active Database simulator to allow nested transaction simulation. From December 1995 to January 2001 (5 years 2 months) Software Design Engineer @ Design of the AIS system for the campus administration. From 1994 to 1995 (1 year) Software Developer (contractor) @ Design of the Point-Of-Sale (POS) system. From July 1990 to June 1992 (2 years) PhD, Computer Science, summa cum laude @ Technische Universität Darmstadt From 1995 to 2001 Master, Computer Science @ National Cheng Kung University From 1992 to 1994 BA, MIS @ National Chengchi University From 1989 to 1992 Ming-Chuan Wu is skilled in: Databases, Distributed Systems, Scalability, Query Optimization, Cloud Computing, Software Design, Microsoft SQL Server, Software Engineering, SQL, Big Data, Indexing, Software Development, Relational Databases, C++, MapReduce, Algorithms, Parallel Computing, Data Warehousing, Object Oriented Design, Agile Methodologies, Processing, Integration, Parallel Databases, Cloud Scale Query...