About Forecast Alpha

Built on evidence, not promises.

Forecast Alpha is a prediction-market research platform. Every signal is timestamped before markets resolve, every model probability is public, and every strategy in the catalog states its evidence stage and blockers before it can be considered for distribution.

We don't publish signals to sell subscriptions. We publish them because a verified, point-in-time track record is the only honest way to tell whether a research process actually works.

What's been built

Research terminal

A probability model across 300+ Kalshi prediction markets. Every read is logged before resolution — the record is public and cannot be retroactively changed.

Execution-evidence engine

The platform records arrival, quote-touch, fill, post-fill, costs, capacity, and broker reconciliation. Broker submission is a separate fail-closed capability and is not granted by a backtest.

Strategy catalog

A curated catalog of strategies labeled by source of return, evidence stage, observed record, and failure mode. Outcome-only, paper, shadow, and limited-live records are never blended.

Strategy distribution controls

The product includes the account, entitlement, risk, and audit controls needed for future strategy distribution. Availability remains subject to evidence maturity, explicit opt-in, and legal review.

CME futures lane

A parallel research lane for CME Micro BTC futures via IBKR, currently in paper validation. Same gate-before-live discipline as the Kalshi lane.

Founder

Brian Lydon

Brian built Forecast Alpha to bring systematic, evidence-based research tooling to prediction markets — a venue where the information efficiency questions are genuinely open and the data pipeline is rich enough to test real hypotheses.

The platform is built with Next.js, Supabase, Prisma, and TypeScript, with guarded Kalshi and IBKR integrations, paper and shadow research lanes, and server-only evidence controls. Research readiness does not imply authority over customer capital.

forecast.alpha.support@gmail.com

Philosophy

Prediction markets are one of the most transparent pricing mechanisms that exist. Prices are probabilities. Outcomes are binary and fast. The edge is measurable in a way that equity or macro markets rarely allow.

The goal is to build infrastructure that makes that edge findable, verifiable, and executable — for individual researchers and eventually for institutional players who want exposure to this asset class without building the plumbing themselves.

Nothing advances without passing its recorded evidence gates. The model versions under review are the same versions whose track record you can audit on the track record page.

Audit the track record How it works Get in touch