I have been working in the field of DS/ML for 3.5 years and am currently based in Espoo, Finland。

I have hands-on experience on the end-to-end DS project lifecycle, from problem identification, data analyzing and experimenting, creating MVP, writing production-ready code to deployment. Building models is important to solving problems, but demonstrating impact and creating value are the final goal.

To achieve that architecture needs to be designed carefully and multiple tools need to be used to ensure a successful delivery. It is challenging and I am excited to deal with any incoming challenges and get my hands on them. Therefore I am really interested in the area of MLOps.

Reflecting on my work, I have made many contributions, either independently or in collaboration with different engineers (e.g. data engineer, backend engineer, Android engineer, SRE). In my previous company N26, I have designed and developed various projects around FCP(financial crime prevention), including fraudster detection, transaction risk rating and initial customer risk score. As a result, the financial loss caused by crime or fraud and operational cost are reduced(by 40%) and the services are more robust.Besides getting satisfying performance from modeling, I have contributed a lot to MLOps, like setting up pipelines and monitoring systems, to ensure excellent practices in scalability, reliability and maintainability. Additionally, I was highly involved in a face verification project which was the first deep learning project of the company. I participated in many meetings to know the goal and create the roadmap and afterwards led the project from the technical side.

Tech stacks:

• Programming/Scripying languages: Python, C++, Java, Matlab, bash.
• DS/ML/visualization related packages: Pandas, Numpy, Jupyter-lab, Scikit-learn, XGBoost, Matplotlib, Seaborn, Plotly and more...
• Deep Learning Frameworks: TensorFlow, PyTorch.
• Monitoring systems/tools: Datadog, Kibana.
• Version control tools: Github, Gitlab.
• ML lifecycle management tool: MLflow.
• Containerization tool: Docker, Docker Compose.
• CI/CD tools: Jenkins, Github Actions, Travis CI, Kubernetes.
• AWS infrastructure: S3, SageMaker, Redshift, EC2, ElastiCache.
• Infrastructure management tool: Terraform.
• BI tool: Metabase.
• Databases: MySQL, PostgresSQL.

Feel Free to email or connect me on and hope we can have a chance to collaborate together.

我在DS/ML领域拥有三年半的工作经验，现居芬兰艾斯堡。

我对人工智能，机器学习，计算机等方向的技术课题很感兴趣，正朝着更深的技术/工程积累迈进。我在金融科技（FinTech）领域拥有参与多个不同主题的数据科学（例如，人脸比对, 用户评分，交易分类等等）项目经验， 并负责数据科学项目端到端的设立（从最初的问题定义，到探索分析，到代码实在，并在满意其效果之后将其部署上线）。

眼下我对MLOPs十分感兴趣。在过往的工作中我意识到，根据待解决的问题和拥有的数据选择并训练出合适的模型是很重要的一部分，但这还不足以创造真正的价值。能够将其部署上线并接入生产环境的流量才是实现价值的开始，而这需要利用不同的工具来实现功能并保证代码的质量。 对我而言，这是一个十分有挑战性和成就感的事情。根据要求设计架构和通过团队合作逐步将其实现十分有意思。同时这也意味着不断的学习，不断吸收新的东西。自己会朝着这条路走下去，并力求能做出更多自己的贡献。

我十分看好技术发展的未来趋势，但同时也坚定地认为怎么样更好地利用技术来创造更大的价值、来给全世界全人类带来福祉，应是发展同时必须考量的问题。希望今后有机会能够投身其中，贡献出自己的一份价值。若有相关机会，请通过下文提及的方式联系我，不胜感激！

作为个体，我们很难控制这个世界的走向，但万幸的是我们能做出选择（或者在争取选择权的路上）。希望放眼今后，我能多做出几个让我不后悔的选择，同样祝愿看到这里的你，能够勇敢地选择你喜欢（哪怕只是一时喜欢）的决定。

技术栈：

• 编程语言：Python, C++, Java, Matlab, bash.
• 数据科学/机器学习/视觉化相关的库： Pandas, Numpy, Jupyter-lab, Scikit-learn, XGBoost, Matplotlib, Seaborn, Plotly 以及更多...
• 深度学习框架：TensorFlow, PyTorch.
• 程序监控系统/工具： Datadog, Kibana.
• 版本控制工具： Github, Gitlab.
• 机器学习生命周期管理工具： MLflow.
• 容器化/开发环境设置工具： Docker, Docker Compose.
• 持续集成/交付工具: Jenkins, Github Actions, Travis CI, Kubernetes.
• 亚马逊云服务基础设施： S3, SageMaker, Redshift, EC2, ElastiCache.
• 云基础设施管理工具：Terraform.
• 商业智能分析工具： Metabase.
• 数据库: MySQL, PostgresSQL.

欢迎通过邮箱和与我联系，期待有认识或者合作的机会！