Data Analyst/Python Developer/Data Scientist with 15+ years of experience (mostly in insurance and banking).
I have a lot of experience in SAS, SQL, various databases, Python (not a full-stack developer), Big Data,
predictive modeling, machine learning, AML, etc.
I have working knowledge of Databricks, Apache Spark, PySpark, Snowflake, Cloud
and AI Agents, among others.
I'm a US citizen, had a green card through National General (NY) previously.
I'm looking for a new exciting role where I can work remotely or 2- days onsite. (Looking for work-life balance,
but doesn't mean I don't work hard!). I can relocate (FL, NC, PA, NY, NJ and TX).
I was promoted to Data Manager at NatGen by the Chief Actuary. I was rehired by Wells Fargo for a second project.
I have great technical skills -- I love working with technology and am a fast learner when it comes to new tools (some jobs
require tools that can be mastered on the job). For example, at AIG I learned how to create ETL's to
convert flat files on an IBM mainframe to SAS tables on Windows, on the job — it uses COBOL types.
My work style is independent, I require minimal supervision. In my last role with Wells Fargo I was a go-to person for challenging technical questions.
I'm open to transitioning back into data science, I'm great with data management and experienced with predictive modeling.
I built a logistic regression model to predict the risk of auto policyholders at an insurer and it was deployed in their underwriting system,
which allowed them to select the best risk and deny coverage to the worst risk, leading to an improved loss ratio.
I also built a GLM model to predict the bare land value of properties in Orange County, CA, using real estate and GIS data (hedonic model).
I'm also open to transitioning into Python development, especially with employers that value a
strong background in math — which helps with algorithm development and handling complex problems.
I'm fine with other roles that match my skillset as well.
I have degrees in both statistics and math from the University of São Paulo (the best in Brazil). I also have an MBA from EAESP-FGV in São Paulo (an accredited business school).
| Accreditor | Country |
|---|---|
| AACSB (Association to Advance Collegiate Schools of Business) | USA |
| AMBA (Association of MBAs) | UK |
| EQUIS (EFMD Quality Improvement System) | Belgium |
A dashboard created with SAS/GRAPH
SAS stored processes to automate data extracts
A Proc Report Excel output (redacted)
I have a knack for programming, both in SAS and in Python and in any of the other languages I've used, such as VBA and Vbscript (it's about algorithm development, not memorization). My background in math helps me with it.
Here are some projects that I've developed (not just in Python):
I'm now working on my first Claude AI agent, a job search workflow hosted here on GitHub.
It uses Python code with a SQLite database and the API for model Claude Sonnet 4-4.6, which has excellent reasoning for tailoring resumes.
Here is an in-progress write-up of the workflow ➔ AI agent
I have finally uploaded my latest paper in Number Theory here.
It introduces a new symmetry relation for the Lerch Φ function that might have implications for proving if a number is algebraic or transcendental.
It is a generalization of the Riemann functional equation with three variables (the Riemann functional equation is the particular case b=0).
Above all, this new paper demonstrates with neat elementary math how a functional equation can be derived -- with the added bonus of an extra variable.
Disclaimer: typos and errors might still exist (it hasn't been peer reviewed yet).
Using GitHub for the first time as an option to pre-print repositories (using Google Analytics to track page hits).
Please note that LaTeX manuscripts are converted to HTML with a tool that is not failproof -- so the pages are not flawless.
Data Analysis
Data Science
Insurance
If you are an Indian recruiter, please only contact me if you are working directly for the client!
Indirect recruiters only make one's job search longer and harder (false leads).
PS - I will block the domain of recruiters that ask for copies of documents or ask inappropriate questions that show lack of understanding
of the US job market/culture. These recruiters could not get you an interview anyway.
Here's a little guide to some iTunes/WMP scripts I've created ➔ Music Scripts
