Houkun Zhu Email & Phone Number

Berlin, De

Berlin, Germany

Berlin, Germany

My main responsibility was to work on the WvSC project with Siemens Energy and conduct research with the focus on cloud computing and artificial intelligence.◆ Designed and implemented Machine Learning algorithms for automating MRO processes on the Siemens shop floor, achieving an accuracy rate of 90%+.◆ Significantly enhanced the performance of distributed file systems by 91.8% through the utilization of a Deep Reinforcement Learning algorithm.◆ Designed and developed an RBAC-enhanced LLM QA system to enable role-based QA to different resources.◆ Maintained both AWS and on-premise cluster infrastructures using Ansible and GitHub Action, ensuring optimal efficiency and robust security.◆ Devised a comprehensive Machine Learning validation framework, specifically aimed at safeguarding the integrity of ML models and reinforcing their ethical application.◆ Successfully led projects milestone involving coordination with 21 companies and institutes, regularly presenting progress and results to stakeholders to ensure transparency and alignment.

Berlin, Germany

I worked in BIFOLD (the Berlin Institute for the Foundations of Learning and Data ). Our research was to estimate the run time of batch data flow jobs.◆ Deployed and maintained a HDFS, Spark and Kubernetes cluster of 45 nodes using Ansible and Helm◆ Designed and implemented an end to end system for Spark job submission, metrics collection and runtime prediction◆ Developed Restful API for different components to communicate◆ Designed and conducted experiments to evaluate our deep learning model for Spark jobs scheduling◆ Published two papers (co-author) from this work

Kiel Area, Germany

◆ Developed a Deep Learning-based recommendation system, leveraging semantic web information to enhance accuracy and relevance◆ Successfully migrated the project from Python 2 to Python 3, ensuring ongoing compatibility and support◆ Implemented and evaluated two different semantic data integration methods (semantic annotation and knowledge graph embedding), allowing for more effective data modeling and analysis◆ Conducted thorough evaluation of the recommendation system across multiple domains, gathering insights and feedback to guide ongoing development and optimization◆ Utilized Github Actions to automate system unit testing and model training, streamlining development and enhancing efficiency

I developed both the backend and frontend of an online survey web app.◆ Interacted with clients to understand their requirements and translated them into actionable items◆ Designed and developed an online survey system using Java, SpringBoot, jQuery, myBatis◆ Automated the application deployment process using a bash script, saving time and reducing errors◆ Worked closely with clients to design a user-friendly interface for the survey system◆ Developed an interactive and real-time administrative dashboard using Plotly to monitor survey results◆ Analyzed survey results and produced interactive reports to communicate findings to clients

Shanghai City, China

◆ Ingested streaming data on 21 diverse primary data source using Java, Spark, Kafka, Memcached, HBase, MongoDB◆ Analyzed news propagation path and key opinion leader using Spark Streaming◆ Automated deployment process using bash script and Jenkins, saving 40 hours of manual work per month◆ Improved performance by 30% on duplicate news data removal by using a new algorithm (MinHash)◆ Developed Restful API using Java, Spring Boot and MyBatis◆ Designed and developed an auto-generated insight newsletter for stakeholders,saving 20 hours of manual work per month◆ Assisted team manager to promote Jenkins for Continuous Integration

Organized students activities and assisted the advisor for students administration. Our class won the Best Class Organization Award for 4 times.

Master Of Science, Computer Science, 1.4 (Range: Highest 1 To Lowest 5)

Bachelor'S Degree, Computer Science And Technology