Skip to content

Latest commit

 

History

History
61 lines (48 loc) · 2.66 KB

resume.md

File metadata and controls

61 lines (48 loc) · 2.66 KB
Raw
layout title permalink
page
Resume
/resume/

Introduction

I am currently resting and looking for my next job. Until recently, I was a software engineer at Zilliz where I lead a small team with four people develop storage for Milvus. In the five years after graduation, I have been engaged in database development.

Work experience

Zilliz, Shanghai
Senior Software Engineer, Team Lead
Dec 2020 - Jan 2022, Milvus v2.0(cloud-native distributed vector database)

  • Built a message queue library based on RocksDB used in the standalone deployment of Milvus.
  • Built the fetch and delete functions which provides session level consistency.
  • Building and governing the Milvus community(membership, automation robots, GitHub management, technical meetings, etc).

Apr 2019 - Mar 2021, Milvus v1.0(shared-storage-based distributed vector database)

  • Investigated the design and implementation of FAISS(ANN library from Facebook).
  • Participated in the architecture design of Milvus.
  • Built the task scheduler, to support dispatch search tasks and indexing tasks to CPU and GPU.
  • Participated in the development of a new algorithm accelerated by both CPU and GPU.
  • Improved the k-selection algorithm on GPU, extending upper limit of TOPK from 1024 to 16384.
  • Led a team of five people to developing the hybrid search function that enables users to execute a similarity search with scalar filtering.
  • Wrote technical articles about Milvus.

Aug 2017 - Apr 2019, Megawise(GPU-accelerated database based on PostgreSQL)

  • Developed the algorithm of SCAN and AGGREGATION, makes Megawise passed the TPC-H benchmark.
  • Built the compilation system of Megawise through CMake, helps Megawise to integrated multiple third-party libraries.

Zerone(Former Zilliz), Jiaxing
Software Engineer
Feb 2017 - Aug 2017, Zdb(prototype of Megawise)

  • Built the JIT execution engine(dynamically convert SQL to CUDA function) which makes the SCAN and AGGREGATION speed up to more than a hundred time than CPU.

Education

South China Institute of Software Engineering,Guangzhou University
BE, Internet of things
Sep 2012 - June 2016

  • Won the first prize in National Undergraduate IoT Design Contest in 2015.
  • Won the second prize in National College Student Smart Internet Innovation Application Design Contest in 2014.
  • Won the second prize in Guangzhou University Programming Contest in 2014.

Skills

  • C++(Advanced)
  • Python(Advanced)
  • CUDA(Advanced)
  • Go(Intermediate)
  • Docker(Advanced)
  • Linux(Advanced)
  • Git(Advanced)

Other

  • Learned CUDA programming and obtain certification from NVIDIA.