Antelope — AI-Powered Polling & Campaign Intelligence

The Problem

The problem at the center of Antelope emerged over a decade of observation across academia, startups, and market research.

In the early 2010s, a pattern became visible in political science and sociology programs: students arrived excited about understanding how societies work, but struggled when they encountered the quantitative tools required to do rigorous analysis. The barrier was not conceptual — it was technical. Tools like R demanded significant overhead, with weeks of learning syntax and wrestling with implementation standing between a researcher and even simple questions about correlation or causation. The result was predictable: inquiry was deferred, simplified, or abandoned entirely.

This is where the conviction behind Antelope took root — the belief that the difficulty of a tool should never be the thing that stops a person from understanding the world.

The Beginning

The founder's own path made the problem personal. As a graduate student at Columbia with undiagnosed ADHD and dyscalculia, the barrier was not a lack of curiosity or capability, but the tools themselves. A single regression could consume an entire term. Simpler tools like SPSS eased the burden but capped the ceiling — descriptive statistics were accessible, but anything deeper remained out of reach.

The same pattern appeared in the classroom at Rutgers. Students who arrived engaged and capable — many from strong traditional backgrounds in political theory — found themselves avoiding quantitative work, not because they could not grasp it, but because the tooling made genuine exploration untenable. The interest was there. The access was not.

Birth of an Entrepreneur

At Oxford, the founder pursued socio-legal studies — rigorous, but largely qualitative, in part because the quantitative path still felt closed. It was during this period, over the COVID lockdowns, that the seed of a company first appeared.

The catalyst was an unlikely one: Reddit's prediction tournaments. These were small forecasting games where users predicted outcomes and earned tokens for accuracy. Though the tokens held no real value, the act of wagering made people reason carefully about their own beliefs. The founder saw something larger in this — a way to map demographics and identity against opinion and instinct, building a dynamic, multidimensional picture of how people actually think.

That vision became Napolleon, a predictive market for business intelligence and the founder's first entrepreneurial venture, accepted into the Oxford University incubator. Napolleon did not ultimately succeed as a product. The environment could only support so much, and the venture wound down. But it was not a wasted chapter — it was the one that pointed toward the real problem.

Pivots and Research

The market research conducted for Napolleon proved more valuable than the product itself. In conversation after conversation, a consistent truth surfaced: every organization used surveys, and almost every organization disliked them. The incumbent tools — nearly two decades old — had grown by accretion, with features bolted onto dated architectures, distribution and analytics siloed, and user experience reflecting accumulated compromise rather than design.

After Napolleon wound down, a period of reflection followed. The survey problem would not let go. An early collaboration with Thomas Petersen briefly explored a hybrid of predictive markets and surveys before simplifying to the core: surveys, but with real analytical teeth — no-code Python, regression, the deeper work that had always been locked behind tool difficulty.

The first instinct was to build for business broadly. It did not work. Pilots generated interest but not sales; businesses defaulted to incumbents without a compelling reason to switch. The breakthrough came only when the focus narrowed to a single market that lived and died by exactly this problem: political campaigns.

Today

Campaigns need to understand their districts — what constituents fear, what they want, what moves them. Yet the tools available to them split into two inadequate halves: outbound engines that broadcast and measure reach, and survey tools with no integrated analytics. The loop was never closed. Listen, analyze, act — each lived in a separate platform, under a separate vendor.

The campaigns that need this most — downballot races, exploratory committees, small PACs, and local organizations — operate on a fraction of national budgets. They needed something fast, affordable, and integrated. The market offered them nothing. Antelope was built to close that gap: survey creation, advanced analytics, geographical intelligence, and outbound messaging in a single platform, priced for the races that matter most but have been systematically underserved.

The Path Forward

Antelope began in 2026 with a clear conviction — that understanding should be accessible, that data should inform strategy rather than replace it, and that the leaders who take the time to truly know their communities deserve tools worthy of that work.

The company owes a quiet debt to those who shaped the journey but cannot be named here: an early co-founder who believed in the vision before departing for academia, collaborators who lent their design and technical insight, and friends who opened doors when the idea was still unproven. Their contributions are woven into what Antelope has become.