Open to opportunities

Abdon Islas

@abdonislas

Message

I build AI and quantitative tools for pricing, risk, portfolios, and financial decisions.

Mexico

Message

What I'm looking for

I'm looking to build AI, machine learning, and quantitative finance products where I can turn financial theory into practical tools for risk, pricing, portfolio, and investment decisions.

At Softtek, I build end-to-end machine learning and AI tools for financial services, including a live small-business credit scoring model handling about 5,000 applications monthly and transaction forecasts that achieved 8.3% MAPE.

I also built portfolio optimization and Monte Carlo tools supporting investment rebalancing decisions, automated reporting from four databases to reduce an eight-hour process to under one hour, and created language-model evaluation frameworks adopted for chatbot quality control. My work combines Python, SQL, quantitative finance, and practical communication with business teams.

I actively trade stocks, futures, and options with my own capital, using volatility, drawdown, and Value-at-Risk to keep research grounded in real market behavior.

Experience

Work history, roles, and key accomplishments

Current

Data Scientist / AI Developer

Current

Softtek

Jan 2022 - Present (4 years 7 months)

Built end-to-end machine learning pipelines for a financial services client, including a live credit scoring model handling about 5,000 applications monthly. Forecasted weekly transaction volume with SARIMA and Prophet, achieving 8.3% MAPE, and automated reporting reducing manual effort from 8 hours to under 1 hour.

Python SQL Machine Learning SARIMA Prophet Mean Variance Optimization A B Testing Prompt Engineering LLM Evaluation AWS Docker Git

Education

Degrees, certifications, and relevant coursework

Instituto Tecnológico y de Estudios Superiores de Occidente (ITESO)

Bachelor of Science, Financial Engineering

B.S. in Financial Engineering with coursework in stochastic calculus, derivatives pricing, fixed income, econometrics, financial risk management, time series, and machine learning. Capstone project involved building portfolios and comparing strategies across a multi-asset universe.