Dr. Randal Olson

I'm building technology that serves humanity through privacy-first AI applications and human-centered design. At Wyrd Studios, I own the AI stack and product strategy, turning messy human workflows into reliable automation that gives people more control over their digital lives.

My approach is pragmatic and human-centered: ship useful automation, measure real outcomes, and keep humans in the driver's seat. I prioritize sustainable value over hype, designing calm technology that reduces cognitive load, cuts toil, and compounds over time.

At Wyrd Studios, we're building subscription-based tools that give users control over their digital lives, with business models that align success with user satisfaction. Our independence frees us from growth-at-all-costs pressures, allowing us to maintain singular focus on solving real problems for our users.

Through RO Consulting, I guided companies navigating the complexities of data science and AI. My work involved dissecting complex problems and architecting custom AI solutions that drove growth and innovation. Highlights included building an AI-automated travel package system for GuideToEurope(opens in new tab) and advising numerous startups on their AI strategy.

Turning cutting-edge AI into real-world business wins. I translated complex client challenges into concrete technical architectures - from LLM-powered conversational systems and computer vision pipelines to predictive analytics platforms and automated decision engines. Across edtech, biotech, finance, and enterprise sectors, my approach identified the critical 20% of business effort that delivered 80% of the value, architecting full-stack solutions directly connected to business KPIs.

As a guide through the rapidly evolving AI landscape, I helped clients separate genuine opportunities from hype while building future-ready technology strategies. My solutions leveraged deep expertise in data science, cloud deployment, and AI engineering to create systems that went beyond showcasing technical capabilities and drove quantifiable ROI through streamlined operations, joyful customer experiences, and truly data-driven decision making.

Data science and AI for EdTech to meet students where they are in their learning journey and improve their learning outcomes.

As a founding member of the AI research team, I helped build ETL systems at the intersection of the wet labs and the AI team, cutting data transfer rates from days to minutes. This data pipeline enabled us to pretrain, fine-tune, and evaluate LLMs for drug discovery. My work on the ETL systems and AI solutions contributed significantly to the filing of one patent(opens in new tab) and remains a cornerstone of Absci's innovative drug discovery pipeline(opens in new tab).

I sat on the FOXO Scientific Advisory Board and advised on the data science and machine learning technology developments at FOXO. This was a part-time contract role.

As the foundational Data Scientist, I embarked on a mission to manifest our company's vision of integrating epigenetic testing into life insurance. Starting from scratch, I built and led a team uniquely equipped to merge cutting-edge epigenetics research with advanced analytics and AI, developing new standards for mortality risk assessment in the industry. Since 2018, my team and I:

• conducted observational research studies with human subjects,
• performed data collection and cleaning on datasets containing over 800,000 molecular measurements and hundreds of human and wellness indicators,
• developed custom DNA methylation processing software from scratch (e.g., https://github.com/FoxoTech/methylprep(opens in new tab)),
• built a cloud-based serverless automated DNA methylation processing platform,
• curated our data into a proprietary research database on the cloud,
• prototyped MVP machine learning models based on our proprietary research database using a state-of-the-art AutoML platform,
• deployed said MVP models into production with a state-of-the-art MLOps platform, and
• integrated said deployed models into a customer-facing web platform.

Issued Patents:

• Olson, Randal S. and Chen, Brian H. Machine learned epigenetic status indicator(opens in new tab). Issued U.S. Patent 11795495.
• Sabes, Jon; Chen, Brian H.; and Olson, Randal S. Machine learning model trained to determine a biochemical state and/or medical condition using DNA epigenetic data(opens in new tab). Issued U.S. Patent 11817214.

As the foundational Data Scientist, I embarked on a mission to manifest our company's vision of integrating epigenetic testing into life insurance. Starting from scratch, I built and led a team uniquely equipped to merge cutting-edge epigenetics research with advanced analytics and AI, developing new standards for mortality risk assessment in the industry.

After working as a postdoctoral researcher for a year, I transitioned into a staff researcher position in Prof. Jason H. Moore's lab at Penn's Institute for Biomedical Informatics. At Penn, I continued developing state-of-the-art machine learning algorithms with a focus on biomedical applications. I also continued aiding Prof. Moore in developing a biomedical visualization laboratory, which will offer machine learning and data analysis as a service to PennMed faculty and physicians. While holding this role, I also assisted the lab in integrating the lab's research projects into the Python data science ecosystem. This entailed translating research code into performant and well-engineered software. Some of those projects include:

• https://github.com/EpistasisLab/scikit-rebate(opens in new tab)
• https://github.com/EpistasisLab/scikit-mdr(opens in new tab)
• https://github.com/EpistasisLab/pmlb(opens in new tab)

I joined Prof. Jason H. Moore's research lab at Penn's Institute for Biomedical Informatics. At Penn, I developed state-of-the-art machine learning algorithms with a focus on biomedical applications. I also aided Prof. Moore in laying the groundwork for a biomedical visualization laboratory that will be used as a collaborative research space by the Penn medical school. While holding this role, I pioneered the field of Automated Machine Learning (AutoML) and built TPOT, one of the most-used open source AutoML libraries in the world. TPOT has since grown into a mature open source library on its own: https://github.com/EpistasisLab/tpot(opens in new tab)

I spent 4 years working on a dual major Ph.D. at Michigan State University working with Prof. Christoph Adami. During my time there, I developed a high-performance C++ platform for digitally studying the evolution of intelligent animal behavior over long, evolutionary timescales. Using this platform, I published several articles in peer-reviewed journals and conference proceedings to share what I learned about the evolution of cooperation and intelligence.

I worked as a civilian employee for the Air Force developing tools for the Electronic Warfare division. My time here was primarily spent as a software engineer working on classified software in a larger software development team.