Experience
Ailys
Machine Learning Data Scientist
Apr 2025 – Jul 2026
As a Machine Learning Data Scientist at Ailys, I worked on ML for insurance underwriting and credit decisions, spanning data foundations, prediction models, and an LLM agent.
I led development of an LLM agent, built prediction models from unlabeled insurance records, and researched reject inference for credit applicants. Through close work with domain experts, I helped turn business questions into verified code and models.
01. KPI Maker
One of my primary focus areas was KPI Maker, helping domain experts turn the KPIs they describe in plain language into working code.
- Led development of KPI Maker, an LLM agent that turns domain experts' plain-language KPI descriptions into verified code, producing correct values for all 12 test KPIs.
02. Claim payout model
I also built a claim payout model for a life insurer, predicting which claims will be paid after a contract is signed.
- Designed prediction targets from unlabeled records by using PMI to link diseases that co-occur in claim histories more often than chance.
- Developed a multi-label claim payout model predicting which of 1,580 diseases and 14 benefit types will be paid within two years after contract signing, reaching 93% recall and 88% specificity.
03. Additional contributions
In addition to modeling, I contributed to the data foundation and credit research behind underwriting decisions.
- Participated in building a data mart for underwriting KPI simulations and analyses by designing aggregate tables and developing preprocessing logic.
- Researched reject inference for credit card applicants without credit history to reduce the bias of default models trained only on approved customers.