Stefano Meschiari, Ph.D.
Data Scientist
I am an experienced Data Scientist with a background in scientific research. As a Data Science leader with more than 8 years’ experience, I use data and algorithms to deliver solutions that are practical, durable, and trustworthy. I bring a scientific mindset, rigorous craft, and people-first attitude to every team I join. I thrive on cross-functional projects that require strong leadership across many teams and quarters.
Skills
- Machine Learning (supervised and unsupervised classification and regression; fraud detection; custom ML algorithms development; high-performance numerical algorithms; applied ML research in the security and fraud space)
- Data and Software Engineering (high-volume ETL and ML pipelines on AWS and Databricks; data products for BI/internal analysis; web-based applications and games; open-source development; Spark/SparkML, H2O, Python, Java, Scala, C, JavaScript, React)
- Data Science and Analysis (causal inference and A/B testing; dashboards, tooling, and reproducible reports for executive decision-making; Python, R, SQL, Streamlit, Shiny, Databricks)
- Team Leadership (mentoring junior teammates to senior positions; scoping complex timelines and deliverables, evaluating risk and impact, and acting as a data advocate across multiple teams; distilling complex concepts to stakeholders and users)
Experience
- Protected our customers from unauthorized access to their merchant accounts. I built ML models, rules, and processes that protect Stripe users accounts, balances, and data. Drafted multi-quarter roadmaps, reported on losses against budget, and delivered status updates to leadership. My work delivered more than $2M/yr in prevented losses and a 2x decrease in accounts flagged.
- Guided our Risk Team through tradeoffs. I designed A/B experiments, analyses, and simulations on tradeoffs between financial losses and customer pain when altering our risk appetite. I provided recommendations for new operating points, financial projections, and dashboards for leadership and our partners. The new operating points resulted in an estimated $3M/year in customer churn reduction.
- Remediated large-scale incidents. As a part of fraud incident response, I worked with urgency to analyze new fraud patterns, stand up incident dashboards, build models to predict terminal losses and plug detection gaps, and conduct post-mortems. I represented Stripe with partners such as Mastercard.
- Built new customer-facing features. I joined a team of Risk PMs, front-end, data engineers, and designers to deliver a significant new merchant feature (Merchant Risk View). I designed the data model and underlying queries, validated the semantics of the fraud metrics, built and stood up a mock implementation feeding from real customer data, and paired with engineering to successfully deliver the feature on a hard marketing deadline.
- Led the research and development of the data science platform that powers Duo Trust Monitor (demo).
- Collaborated with Product and Design teams to seamlessly integrate machine learning and effective data presentation into Duo’s product, prioritizing usability and customer trust. Worked closely with customers to understand their use cases and validate new capabilities.
- Coordinated Data Science and Engineering work to ensure alignment, manage expectations, and hit delivery milestones.
- Led team through growth from 1 to 10 IC members, navigating substantial organizational change. Mentored 5 peers in data science and engineering. Served as interim manager for multiple data teams as required. Developed technical interviews on research, algorithms, and ML architecture.
- Researched supervised and unsupervised algorithms tailored to the security domain, focusing on simplicity, explainability, and scalability.
- Worked with domain experts to translate cybersecurity knowledge into expert rules and heuristic layers that complement ML models.
- Developed ETL and ML infrastructure to process data, build models, and surface threats and authentication anomalies at scale. Duo processes almost a billion authentications every month.
- Created reports, dashboards, and prototypes that guided technical and UX decisions. Delivered presentations, demos, and trainings internally (for engineering, sales, and VP-level executives) and externally.
- Submitted 2 patents to USPTO, including a first-inventor patent for our proprietary ML algorithm for threat detection.
- Created and improved on machine learning pipelines and tooling to model university student outcomes. Improved the custom modeling platform, reducing batch training and scoring running time and cloud costs by half.
- Developed new ML models and custom classification algorithms (in R/ Caret, SparkML, and JavaScript) and back-end data APIs for internal services.
- Developed applications that surfaced institution-level insights and visualizations, empowering Sales and executives with timely data- driven talking points and facilitating new partnerships and upsells.
W. J. McDonald Postdoctoral Fellow
2012-2016
SAVE/Point, Principal Investigator
2014-2016
As an astronomer, I divided my time between conducting research in theoretical astrophysics and exoplanet detection, building libraries tooling for the astronomical community, and devising apps and games for astronomy education.
- Led the data analysis effort for the Lick-Carnegie science collaboration (~20 scientists across the United States). Analyzed high-value time series data captured with Keck, APF and Lick telescopes using my Markov-Chain Monte Carlo code, Systemic. Systemic has been used to discover more than 40 new planetary systems.
- Wrote high-performance, parallelized codes to solve ordinary and partial differential equations modeling planet formation.
- Developed Super Planet Crash, an HTML5/JS game that was played more than 15 million times and was covered by The Verge, IO9, Huffington Post, and others; and Systemic Live, an HTML5/JS web app used at Caltech, UF, UT, MIT, SJSU, UD, Yale, Columbia, Coursera MOOC to teach students about data analysis and modeling.
Education
- 2012 – Doctor of Philosophy (Astronomy & Astrophysics), University of California at Santa Cruz. Received Whitford Prize for highest achievement in research, coursework, and teaching.
- 2006 – Master of Science (Astronomy, with highest honors), University of Bologna
- 2004 – Bachelor of Science (Astronomy, with highest honors), University of Bologna
Publications, Academic Honors and Awards
- Published 8 first-author refereed publications on time series analysis, optimization, and physical simulations (cited 437 times); a total of 17 refereed publications (cited 1,444 times). See research page.
- 2014 – Meschiari, S. (PI), Ludwig, R., Green, J., Interactive Education Tools in the Public Square (Award: $2,800, for creating an interactive outreach experience); Bringing the Tools of Research Direct to the UT Classroom: Systemic, a Virtual Lab for Students (Award: $87,710)
- 2010 – Award for Excellence in Teaching
- 2008 – Whitford Prize for graduate academic performance
- 2006 – Regents’ Fellow, University of California