Desktop required

This experience is optimized for larger screens. Please visit on a laptop or desktop.

Where markets meet the singularity.

Autonomous AI agents that turn compute into alpha.

Event Horizon Labs 2026
We operate a platform that automates quantitative research —
generating hypotheses, backtesting strategies, and analyzing results.
Hundreds of agents run in parallel
and learn from what the others discovered.
We are building a fund that scales
natively with artificial intelligence.
We're hiring founding engineers →
PLATFORM

Turning compute into capital

An orchestration layer dispatches parallel agents, routes results through markets, and compounds what they learn. Every cycle makes the next one smarter.

Alpha Research — Orchestrator 8 AGENTS ACTIVE
EXPERIMENTS 38,471
UNIQUE STRATEGIES 12,041
BEST SORTINO 4.84
PORTFOLIO SORTINO 6.12
SESSIONS 142,910
THE RESEARCH LOOP

Agents that research, decide, and improve.

Each session accesses the shared knowledge base, synthesizes prior findings into novel hypotheses, backtests rigorously, and commits what it learned. The next session starts smarter. Zero human intervention.

ORIENT
14,291 session logs · 12 lesson files · target: strategy_d4c71e8
Loaded an equity momentum strategy producing a Sortino of 1.22. Prior sessions identified horizon spacing as the binding constraint — lookback windows too coarse in volatile regimes. This session will attack it.
HYPOTHESIZE
4 candidate strategies · 2 new features designed
A: tighter geometric spacing. B: adaptive spacing weighted by volatility. C: ultra-short regime filter. D: PCA-residualized decomposition. Each paired with a multi-horizon vol ratio and order-flow imbalance z-score.
EVALUATE
4 screened → 2 promoted to full backtest
C and D failed drawdown and IC gates on coarsened 6-month data. A and B advanced to 3-year backtest with regime-conditional splits.
candidate_A SR 3.18 · DD 2.7% · Calmar 31.4 PASS
candidate_B SR 0.9 · DD 2.6% · Calmar 8.1 FAIL
SYNTHESIZE
Sortino 1.22 → 3.18 · committed as strategy_d4c71e8_v2
Tighter geometric spacing resolved the horizon problem. Wrote lessons on horizon ratio sensitivity and feature pairing constraints.
6 of 7 research tasks closed — horizon spacing resolved. Next session will target the final open task: cross-asset correlation decay.

Incumbents scale headcount.
We scale agents.

AI labs don't trade. Quant firms don't build agent infrastructure.
EHL occupies the intersection — an agentic platform that starts where the economics are best.

Team from Citadel, Jump Trading, Stanford, Caltech, Berkeley
INVESTORS
WE'RE HIRING

Founding roles for researchers and engineers.