Summary
Overview
Work history
Education
Skills
Websites
Languages
Affiliations
Timeline
Generic

Adrian Lei

City of London

Summary

Curious, self-driven, and ambitious Data Scientist with 7 years of experience in the fintech/insurtech industry; having experience in credit risk modelling, product pricing strategy, and portfolio risk management. Skilled at solving technical machine learning problems with the business problem in mind. Has strong experience in communicating with non-technical personnel to drive decision-making. An avid learner who is always eager to tackle the next challenge.

Overview

7
7
years of professional experience

Work history

Credit Data Scientist

Starling Bank
London
09.2022 - Current
  • Built Python and SQL models for internal and external regulatory reporting purposes and orchestrated thedeployment of these models through DBT
  • Built Looker dashboards to track arrears behaviour on overdraft loans that assist collections teams in their roles
  • Developing a machine learning model that predicts the risk of a customer going into a pre-arrears process with the aim of implementing an early intervention strategy with the customer success team. Using kub
  • Developing a customer segmentation model that aims to identify 'main account' customers, working with a diverse set of stakeholders including marketing, regulatory and treasury teams to define outputs and uses of the model

Data Scientist

Insurami
London
09.2019 - 07.2022
  • Responsible for end-to-end training, cataloguing, and deployment of a risk scorecard based on a dataset of 30M+ sets of financial accounts; utilising XGBoost/LightGBM and developing custom loss functions for use in the model.
  • Implemented pricing methodology and strategy for a first-mover product, building a stress test pipeline to forecast loss ratio outcomes over 5 years.
  • Structured data in SQL databases to build a suite of live risk monitoring tools using Python, SQL, and Mode analytics; utilised these tools to lead monthly risk monitoring meetings.
  • Conducted case studies on 4000+ commercial leases for 20 landlords, including COVID-19 resiliency analysis and generation of early default indicators for aid in portfolio lease management.
  • Presented risk modelling and pricing methodology to various non-technical personnel, including partners and potential investors.

Data Scientist

Liberis
London
10.2018 - 09.2019
  • Responsible for credit scorecard maintenance and performance analysis, handling monthly risk monitoring meetings, whilst proposing and driving improvements.
  • Undertook a scorecard improvement project utilising geo-location data and string fuzzy matching to improve the data matching rate of potential customers; leading to a 20% increase in the population of customers that could be automatically underwritten.
  • Worked with Open Banking data to develop and deploy a transaction classification algorithm via Azure which was utilised in the customer application journey to automate the reporting of card volumes, speeding up the customer application journey.

Volunteer

Ignite Hubs
London
11.2023 - Current
  • Assisted in Scratch/Python coding lessons by guiding students on self-driven projects as well as aiding in preparation of class materials

Education

Bachelor of Arts - Economics & Management

University of Oxford
Oxford
07.2018

Skills

  • Python/SQL
  • AWS/GCP/Azure
  • DBT/PySpark
  • Kedro/Kubeflow
  • Scikit-Learn, XGBoost, LightGBM, Keras
  • Looker, Power BI, Dash

Languages

English
Native
Chinese (Cantonese)
Native
Chinese (Mandarin)
Fluent

Affiliations

  • Close up Magic - worked as a freelance magician
  • Video Games - team leader of a team ranked top 0.1% in Europe for Apex Legends
  • Fermentation - making my own kimchi
  • Sports - basketball & badminton

Timeline

Volunteer

Ignite Hubs
11.2023 - Current

Credit Data Scientist

Starling Bank
09.2022 - Current

Data Scientist

Insurami
09.2019 - 07.2022

Data Scientist

Liberis
10.2018 - 09.2019

Bachelor of Arts - Economics & Management

University of Oxford
Adrian Lei