Institutional-grade equity research on every company in the ASX 200.
Built by an analyst. Powered by AI. Scored in public.
Model portfolio figures, before costs. A young record, labelled honestly.
A familiar experience for many
Most investors go looking for help. Here is what they find.
Slick marketing and aggressive upsells obscure the competent providers from sight.
Some providers run funds. One hand advises you, the other trades against you.
Sell-side research is conflicted by design. Corporate relationships determine what gets published.
This is where we come in.
Every report. Same intensity.
Examine
Every company begins with a full review of financial statements, earnings releases, and regulatory disclosures. We read what the company says about itself, before forming any view.
If the estimates are not precise and defensible, the valuation is worthless. The process exists to make them defensible.
Each company runs through an analysis process covering 100+ pages, distilled into a report you can read in minutes.
The analyst behind the platform
Anyone can ask AI for a report. That is not what this is.
Producing an investment analysis with AI is something many people can do now. Producing analysis reliable enough to build a portfolio from is a different problem entirely. That is the one we set out to solve.
One analyst could not compete with incumbent research operations alone. AI put us on a fairer footing.
No Conflicts
We do not manage money, do not take corporate advisory fees, and do not broker trades. So that you may understand your investments without deceit.
Bloomberg Data
Institutional-grade market data. Not scraped APIs or delayed feeds. We don't settle for anything less.
Framework, Not Opinions
Every valuation follows a structured, repeatable process. Our emotions are disengaged from the analysis. Views change when the data changes, not when sentiment shifts.
We score every call we publish from its analysis date, hits and misses alike.
Outside of the data source and the AI models, we depend on nothing and no one.
This platform utilises technology that is still in its early stages of adoption. It will not be perfect, and it will not come without faults. Some reports will contain imperfections. Some views will be wrong. This is true of every research provider and AI platform, whether they would be willing to admit it or not. The difference with us is transparency. When something goes wrong, we do not shy away. We address it. When an investment view changes, you are told why. There are no black boxes here. Everything that is produced can be explained upon request. That accountability is both in the interest of our members and in my own.
I held myself to a very high standard when I started Alpha Insights. AI adoption for me was not driven by curiosity. It was driven by necessity. Even now, every two to three months, what I thought was the ceiling gets reset.
Ryan Lim, FounderJudge the process on its record, not our marketing.
Top 15 long ideas, conviction-weighted. 14 October 2025 to 3 September 2026. Benchmark: S&P/ASX 200 (+1.4%).
Model portfolio figures, before costs. A young record, labelled honestly.
Past performance is not indicative of future results. Indexed daily NAV, before fees and costs; not an actual trading account. From 10 April 2026 the record is the live CORE model portfolio; before that, the long book of the same research signals (linked record).
Don't take our word for it
Get in touch
Have a question, or want a specific company covered? Reach out directly. You'll reach the analyst, not a support desk.