Pengcheng Liu
Analytics Engineer @SumUp
Signup · Get unlimited contacts
WORK HISTORY
Analytics Engineer @SumUp
Berlin, DE
own 100% data modelling work of Lending Unit that offers installment loans (Merchant Cash Advances, Pix Credit, Micro Loans, etc.) to SumUp merchants- champion self-service data analytics - partnering with business and product teams to improve our data literacy and decision making- team up with platform and engineers to build infrastructure for data modelling from scratch. Stack: dbt, Snowflake, Python, Kafka, Airflow, Fleet- do data modelling in an engineering manner: develop, test, debug, document, version control and deploy data models- do data modelling to meet a variety of business needs: credit scoring, origination, customer relationship management, conversion optimisation, accounting, P&L, credit risk management, servicing, collection, etc- create Tableau dashboards on top of data models so as to gain and share insights- create AI Agent that enables text-to-SQL and text-to-analysis with a Web UI. Stack: Docker, Fleet, Streamlit, Langchain, Python
EDUCATION
Humboldt-Universität zu Berlin
Master's degree, Economics and Management Science (Information Systems as focus)
ABOUT PENGCHENG LIU
An overview of my career in the field of Data:In solarisBank: 1) develop, maintain and improve a variety of data products (services, views, reports, dashboards, etc.) to meet stakeholders\' requests and support data-driven decision makings. 2) cooperate with engineers to develop and promote Data Platform as a central place that enables non-Data colleagues to do data storage, data sharing, data visualisation, reporting, modelling, etc. 3) improve unit\'s commercial performance by reducing cost (e.g. build alerts on credit risk) and by spotting new opportunities (e.g. identify potential customers, propose approaches to increase payout volumes)In Babbel: develop Machine Learning Models in Python to predict customer churn and customer lifetime value, these 2 KPIs are critical to subscription-based business (e.g. Netflix, Spotify, etc.).In Recare Solutions: 1) develop backend services in Golang to compute, monitor, export and visualize KPIs from data via ORM; 2) make use of AWS, SQL and Domo to provide BI products to decision makers.In Simby (Another AI): 1) develop Android prototype APP in JAVA to collect and persist hardware data. 2) build ML models in Python (Numpy+Pandas+sklearn) to predict and analyse hardware signals.In HU Berlin: complete my master thesis in Machine Learning: build Ensemble Learning Models (XGBOOST + Linear Models) to predict overdue or default of online consumer loans, in order to deal with the unbalanced dataset problem, which is a common issue for credit scoring.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.