Zoa Research

Historically, quantitative models are domain specific. Brilliant people spend their best years testing features, tuning hyperparameters, and iterating architectures within a narrow domain. But scale is the panacea: large models will find patterns people, and specialized models, could not.

Forecasting generalizes. Zoa trains cross-domain event forecasting engines.

*Automating Iteration*

LLMs:embedded in multi-agent optimization loops and evaluated against fixed policies:can automate the build-test-improve modeling cycle. Think AlphaEvolve for forecasting problems.

*Sample-Efficient General Models*

Today’s forecasting models are narrowly crafted with deep human priors. But larger models will outperform state-of-the-art specialized models.

Unlike existing event models, our models leverage data from across contexts and rely less on human intuition. And compared to LLMs, our models are built with more inductive priors and rely more heavily on inference-time compute:improving sample efficiency.

*Why It Matters*

In the real economy, our models could be useful for forecasting supply chain volatility, energy supply and demand, even earthquake risk.

Science is, Ian Hacking writes, the taming of chance. It is the process of iteratively updating priors (something like: identify uncertainty, conceive experiment to reduce uncertainty, execute, update). If science is uncertainty-reduction, forecasting is a critical measure of progress.

Better forecasting improves our ability to select interesting experiments (roughly those with greatest expected uncertainty reduction) and update priors. Our models will be used by labs and academics in data-heavy domains.

Sam's ex-girlfriend introduced him to Greg back at Carnegie Mellon in 2017, and while that relationship didn't last, their friendship has. After college, Greg went to Harvard Law School, while Sam worked for three years at Jane Street on their Options desk, building & leading a satellite dev team.

High-Level Overview

Zoa Research is a cutting-edge AI company specializing in powerful quantitative forecasting models that generalize across multiple domains rather than being narrowly specialized. Founded in 2024, Zoa builds advanced predictive engines that leverage cross-context data and rely less on human intuition compared to traditional models or large language models (LLMs). Their technology aims to improve forecasting accuracy in complex real-world scenarios such as supply chain volatility, energy markets, and even earthquake risk. This capability is particularly valuable for labs and academics in data-heavy fields, enabling better decision-making and scientific discovery acceleration[2][5].

For an investment firm perspective, Zoa Research’s mission centers on advancing forecasting science through scalable AI models that reduce uncertainty and improve experimental selection. Their investment philosophy likely emphasizes deep technical innovation and cross-domain applicability. Key sectors include AI, data science, and quantitative finance. Their impact on the startup ecosystem is notable in pioneering generalizable forecasting technology that could transform multiple industries reliant on predictive analytics.

From a portfolio company standpoint, Zoa builds quantitative forecasting models serving researchers, traders, and scientific labs who need robust event prediction tools. They solve the problem of limited domain-specific forecasting by creating models that detect patterns across diverse datasets, improving sample efficiency and inference. Their growth momentum is reflected in their rapid formation in 2024, participation in Y Combinator’s Summer 2024 batch, and a focused team of experts from Carnegie Mellon, Harvard, and Jane Street backgrounds[2][7].

Origin Story

Zoa Research was founded in 2024 by Sam Damashek and Greg Volynsky, who met through a personal connection at Carnegie Mellon University in 2017. Although their initial relationship ended, their friendship endured and led to collaboration. Sam brought experience from three years at Jane Street’s Options desk, where he led a satellite development team, while Greg pursued Harvard Law School. Their combined expertise in quantitative finance, law, and computer science shaped Zoa’s vision to build scalable, cross-domain forecasting models that improve upon existing domain-specific approaches[2][7].

The idea emerged from recognizing the limitations of traditional forecasting models that rely heavily on human intuition and domain-specific tuning. By leveraging large-scale models with inductive priors and inference-time compute, Zoa aims to generalize forecasting across contexts, a pivotal innovation in predictive modeling. Early traction includes acceptance into Y Combinator’s Summer 2024 batch and assembling a small but highly skilled team focused on AI and machine learning[2].

Core Differentiators

• Cross-domain forecasting: Unlike traditional models limited to narrow domains, Zoa’s models generalize across multiple contexts, enabling broader applicability.
• Reduced reliance on human intuition: Their models leverage data-driven inference with strong inductive priors, improving sample efficiency compared to LLMs.
• Expert team: Founders and team members have backgrounds from Carnegie Mellon, Harvard, and Jane Street, combining expertise in AI, quantitative finance, and law.
• Focus on scientific discovery: Their models are designed to accelerate experimental design and uncertainty reduction in scientific research.
• Integration of LLMs and traditional ML: They combine large language models with classical machine learning to systematize event-driven trading and forecasting[2][4][7].

Role in the Broader Tech Landscape

Zoa Research rides the trend of AI-driven quantitative forecasting and the increasing demand for scalable, generalizable predictive models. The timing is critical as industries face growing complexity and volatility—such as supply chains, energy markets, and scientific research—where traditional forecasting struggles. Market forces favor AI models that can integrate diverse data sources and reduce uncertainty more efficiently.

By advancing forecasting science, Zoa influences the broader ecosystem by enabling better decision-making and accelerating scientific progress. Their approach aligns with the growing emphasis on data-driven strategies and AI augmentation in finance, research, and risk management. This positions Zoa as a key player in the evolution from domain-specific to universal forecasting models[2][5].

Quick Take & Future Outlook

Looking ahead, Zoa Research is poised to expand its impact by refining its cross-domain forecasting engines and broadening applications in trading, scientific discovery, and risk assessment. Trends such as increased AI adoption, demand for real-time predictive analytics, and integration of machine learning with domain expertise will shape their journey.

Their influence may evolve from a niche AI startup to a foundational technology provider for multiple industries reliant on forecasting. Continued growth will likely involve scaling their team, deepening partnerships with academic and industry labs, and enhancing model capabilities to tackle increasingly complex prediction challenges.

Zoa’s mission to tame uncertainty through scalable forecasting models ties back to the core scientific principle of iterative uncertainty reduction, positioning them at the forefront of AI-driven decision science[2][4][5].