Welcome! I'm Asad Mehasi, a dedicated Data Scientist with a strong foundation in Mathematics and Numerical Analysis. My passion for automation and machine learning drives me to create innovative solutions that add value to businesses.

Learn about me

In addition to my extensive experience in R, Python, SAS, Spark, SQL, Power BI, and Tableau, I’ve made impactful contributions at companies like Swinton Group (part of Atlanta Group), where I integrated Python models into Radar via PMML and automated data preprocessing to streamline workflows.

At Ticker, I worked on pricing a new telematics insurance product tailored for Older Drivers aged 65+. This initiative tackled the challenge of higher premiums often faced by this demographic. While older drivers may have slightly higher claim frequency, their claim severity tends to be lower. Our goal was to develop a pricing strategy that didn’t penalize age alone—leveraging telematics to ensure fairness while managing risk effectively.

In 2024, I took on freelance projects, including the development of a PDF Table Extraction Toolkit. This lightweight tool showcases my skills in OCR, clustering (DBSCAN), and data preprocessing, and is designed to extract tabular data from PDFs—even on machines without GPU support.

I’ve also been exploring how modern Data Science tools like Bayesian A/B testing and DevOps pipelines can bridge the gap in traditional insurance pricing. Historically dominated by actuarial teams, pricing is evolving—and data science is well-positioned to lead that transformation. Organizations that embrace this shift will gain a competitive edge.

As a father, I prioritize balance and bring that same focus and dedication to my professional work. I thrive on collaboration, continuous learning, and solving complex problems. Let’s connect and build something impactful together.

Experimentation

Ongoing explorations and technical development areas I'm experimenting with.

DevOps for Data Science

Integrating Python models into CI/CD pipelines using GitHub Actions and Docker. Exploring deployment of models in production environments with automated testing and version control.

Bayesian A/B Testing

Comparing Bayesian vs Frequentist methods for pricing experiments. Applying probabilistic programming with PyMC to model uncertainty and update beliefs in real time.

Natural Language to SQL

Using open-source LLMs to translate user queries into SQL for insurance databases. Testing fuzzy logic and semantic parsing to improve result accuracy and user interaction.

Portfolio

Models coming soon(ish)! Between debugging code and bedtime stories, family wins the time-slot battle—for now. But don’t worry, the models will make their grand entrance once the little ones are asleep and the coffee kicks in.

In the meantime, here are some of my early Data Science projects from back when I was just getting started:

CLICK TO SEE A PROJECT USING BIG DATA FOR AMAZON INTERVIEW

Linear Regression

Predict yearly amount spent on an online store depending on some numerical features.

Logistic Regression

Predict whether or not a user will click on an ad based off the features of that user.

Open to suggestions

“The goal is to turn data into information, and information into insight.”

Send me an email

Click to send email