Sports - Rock Climbing
Kevin De Jesus
Data Science Manager
Miami,FL
Summary
Meticulous Data Scientist/Engineer who is well versed on taking up multiple roles within an organization. Work with large teams to solve complex problems and have had success in meeting targets that improved business functions. Interested in research but also in delivering timely optimizations in fast moving sectors.
Overview
5
5
years of professional experience
4
4
years of post-secondary education
Work History
Data Science Manager
Idea Financial
01.2024 - Current
- Manage a team of 4 in-house data scientists to solve cutting edge problems within the fintech lending space. Work closely with them to build risk models and to develop and integrate APIs that enhance our automated decision engine. Also manage a team of 3 remote data scientists in Poland that are responsible for maintaining our ETL pipelines to empower the reporting we do through Power BI.
- Engaged in multiple capital raising rounds in order to secure debt facilities to support financing operations. We managed to add 70MM in additional facilities in 2024 to build our loan portfolio and I aided multiple special purpose vehicles by providing them with internal performance data as well as forecasting projections in order to secure funding.
- Develop and oversee risk model implementation and tracking, and automated risk engine enhancement. We have scaled our application throughput and total funding volume 30% year over year while managing to halve defaulted dollar rates since our all time of high of 9% in 2019.
Machine Learning Consultant
Self-employeed
05.2019 - Current
- Help healthcare intelligence companies based in Latin America identify key market trends for hospital equipment.
- Built multiple prospecting models to help product suppliers identify hospitals that would need specific types of equipment using logistic regression and random forest models.
Lead Data Scientist
Idea Financial
01.2020 - 12.2023
- Fullstack data scientist building credit risk pipelines with Python, R and AWS. Use AWS services like EC2 and Sagemaker to ETL, fine-tune, feature engineer and serve models for multiple business cases. Our models are mainly geared for credit risk applications such as preventing losses and saving money on underwriting costs. I have served multiple models including Logistic Regression, Random Forest and XGBoost which have priced and funded thousands of loans. Contributed to the entire cycle from data prepping to model hosting through endpoints
- Maintain automated decision engine which underwrites applicants and provides loan pricing. Heavily increased application processing volume which has increased human productivity by filtering out bad applications and pricing loans through empirically validated methods.
- Build other operational models such as image parsing models with Tensorflow to parse documents such as tax returns and bank statements and use automated LLM chat models to send collection emails to merchants.
- Manage multiple BI tools to monitor model performance and to communicate business KPIs with executives.
- ETL Salesforce data through api integrations and manage database schemas that capture and transform entities in Salesforce that we use for modeling and analysis.
- Maintain relationships with vendors and negotiate pricing to make sure we are getting increased efficiency from our data providers.
Data Science Intern
Idea Financial
07.2019 - 12.2019
- Built credit risk models to mitigate default risk and helped code automated risk engine to underwrite applicants. Reported on strategic insights and crucial KPIs using BI tools such as R Shiny. Capstone project was to build a new loan pricing model using probability of default to determine interest rates, term lengths, and offer amounts. Had to integrate multiple data sources from different credit providers through api in order to get live model predictions. This model ended up replacing the hard cut scorecard method that was used before and aided in bringing the default rate of the portfolio from 9 to 3.5%
Education
Bachelor of Science - Computer Science
Florida Intl. University
Miami, FL 01.2016 - 01.2020
Skills
Machine learning
Statistical analysis
Python
R
AWS
Prompt Engineering
PostgreSQL
Git
Apache Superset
OpenAI - GPT
Docker
Data Mining
Multitasking Abilities
Analytical and Critical Thinking
Accomplishments
- Won first prize in an engineering contest to construct a mobile catapult that could be configured to hit any target given to us at random on a wall 10 meters away with 9 different targets on it
- My team's catapult was the only one with 100% accuracy.10 Data Analysis Project
- Won 1st place at FIU in the AIS Analytics Challenge and then won first place in the National AIS Conference
- Analytics challenge at Temple University
- My team analyzed data from the Open Payments website to determine where the money in healthcare was flowing, potential over/under funding, and which disease groups were being overlooked
- Link to article: https://biznews.fiu.edu/2019/04/fiu-students-capture-top-awards-in-business-analytics-and-blockchain-at-national-technology-conference/
Interests
Chess
Timeline
Data Science Manager
Idea Financial
01.2024 - Current
Lead Data Scientist
Idea Financial
01.2020 - 12.2023
Data Science Intern
Idea Financial
07.2019 - 12.2019
Machine Learning Consultant
Self-employeed
05.2019 - Current
Bachelor of Science - Computer Science
Florida Intl. University
01.2016 - 01.2020