About

Built by experienced investors and traders, for serious self-directed ones.

Catalyst Ahead is built and operated by a team of working investors and traders with years of self-directed-equities experience across the ASX, TSX and US markets. We were tired of refreshing HotCopper, sketching half-remembered guidance dates on whiteboards, and missing obvious read-throughs because the catalyst list lived in someone else's head. So we built the tool we wanted — a platform powered by AI but supervised, end-to-end, by humans who actually trade these names.

Origin

A whiteboard problem.

In late 2025 our small group of self-directed equity investors realised we were spending more time tracking upcoming catalysts than analysing them. Every Monday stand-up looked the same: someone would update a whiteboard with new dates extracted from announcements that had dropped over the weekend. By Friday it was already wrong.

The problem wasn't a lack of data — the ASX publishes every announcement publicly. The problem was extraction. A 14-page drilling update buries the line that matters in paragraph six. A clinical trial briefing hides the readout date in a footnote. No human keeps up across more than ten names.

Then AI got good. Suddenly a model could read every filing, find the forward-looking commitment, normalise it to a date, and grade its own confidence. Catalyst Ahead is what that whiteboard becomes when an AI does the reading.

AI + humans

Powered by AI. Supervised by people.

The Catalyst Ahead platform is powered by AI — advanced language models do the reading and structured extraction across thousands of filings every day. But every prompt, every JSON schema, every guardrail, every edge-case adjustment is written, reviewed, and continuously monitored by experienced investors and traders on our team. Where the model is uncertain, we either tighten the schema, add a manual rule, or flag the output to a human queue before it lands on a customer dashboard.

Concretely, our humans run the playbook on:

  • Context fields — what facts each model is told to look for, in which sections, with which definitions.
  • Data architecture — how extracted records are normalised, deduped, versioned, and surfaced.
  • Guardrails — schema-strict outputs, confidence grading, sanity checks, and "if unsure, return null" rules.
  • Manual inputs — hand-curated supplements (e.g. mines GA's dataset misses), industry-specific overrides, and post-extraction QA.

The result is a tailored response, not a generic AI dump — and a platform that earns the right to sit beside the trader as a research surface, not a replacement for their judgement.

Principles

How we build.

01

Source attribution, always

Every catalyst on the timeline links to the original company announcement. No black-box "trust us" predictions.

02

Confidence, not certainty

We grade every projection High / Medium / Low and show our reasoning. Useful tools acknowledge their limits.

03

AI assists, humans supervise

Extraction is AI-powered, but every prompt, schema, guardrail, and edge-case adjustment is written and reviewed by experienced investors and traders on our team. The AI does the reading; humans run the playbook.

04

No financial advice, ever

We're a data-visualization surface. We surface signals; you decide what to do with them. We will never recommend trades — that's not what this platform exists to do.

05

Pay for what you watch

No bait-and-switch tier games. Plans scale with watchlist size — the only cost driver that actually matters.

2,300+

ASX-listed companies in scope

24/7

Extraction pipeline

6h

Free-tier refresh cadence

~70%

Cost saved via prompt caching

Try it before you commit.

Every market-sensitive date from every filing on your watchlist, on one timeline.

Made in Australia

Product

Company

Legal

© 2026 Catalyst Ahead

Not financial advice · Research only

Not financial advice.Catalyst Ahead is an AI-driven research tool. All projections are inferences from public company announcements. Always do your own research and consult a licensed adviser before trading.