The problem was never information.
A retail investor in 2026 has more data than a professional had in 2006 — free real-time quotes, filings, transcripts, options chains, and a model that will summarize any of it on request. Portfolios still drift, winners still grow unmanaged, and losses still go unharvested until December. The bottleneck moved. It's the decision, not the data.
Three things nothing else quite does.
Nothing knows your book
A chatbot will tell you whether META looks expensive. It can't tell you that your META position is now 2.8× your average allocation, or that you're carrying $4.6k of harvestable losses two rows down, because it has never seen your positions.
You have to know what to ask
Screeners and chat both wait for a query. That works when you already know the question — and fails exactly when it matters most, because the decisions that hurt are the ones you didn't think to ask about.
One answer, no working
Most AI finance products hand down a verdict and hope you trust it. You can't audit a verdict. If you can't see why, you can't tell a good call from a confident one.
Four commitments the product actually keeps.
Reasoning is the product
The recommendation is the cheap part. What makes it usable is the paragraph explaining that a stock re-rated to 26× forward earnings rather than growing into its multiple — because that's the sentence you either agree with or don't. Erina is built so you can disagree with it precisely.
Deterministic where it should be
Not everything should go through a model. Flagging winners past 20%, computing harvestable losses, or listing what reports in the next two weeks are arithmetic — so they're done deterministically, they're always on, and they cost you nothing. The model is reserved for the part that needs judgement.
You always have a choice
Every workflow ends in two or three structures with their numbers attached, and one of them is nearly always to do nothing. A tool that only ever tells you to act is a tool with an incentive problem.
It cannot trade
Erina isn't connected to a broker and has no execution path. Picking an option saves it as a tracked recommendation you can follow, and the order stays yours to place wherever you already trade. That's a design decision, not a missing feature.
An answer you can argue with.
Real output from the sample book. Read the second paragraph — that's the test. If a tool can't make that argument, it isn't reading your position, it's describing a stock.
Should I book, protect, or add?
Positions up 20% or more. Erina explains the reasoning behind each option.
META is your biggest winner — up 44.6% and now the second-largest position in this book. The run is being carried by ad-price growth and Reels monetization, both structural rather than one-quarter tailwinds. That said, at $578 the stock has re-rated to ~26× forward earnings, its widest multiple in three years, and the position is now ~2.8× your average non-mega-cap allocation.
The question isn't whether the story is intact — it is — but whether the position size still reflects your original conviction. If you'd started fresh today, would you buy 60 shares at $578? Most books say no.
Books gains, keeps 45 shares of exposure, and re-anchors position size closer to your average.
Sell a Sep $620 call, buy a Sep $540 put. Downside defined, no shares sold, position keeps compounding.
Revenue and margin rate-of-change still improving. Cutting on multiple alone leaves money on the table.
Illustrative output on sample data. Not a recommendation and not investment advice.
Who this is actually for.
A good fit if…
- You manage your own portfolio and make your own calls.
- You hold real positions and want help deciding what to do with them, not stock tips.
- You want the reasoning so you can disagree with it.
- You'd rather have three options with numbers than one verdict.
- You use options, or you're willing to — collars and covered calls come up often.
Not a fit if…
- You want someone to manage the money for you — Erina isn't an adviser and doesn't take discretion.
- You want automated execution or a trading bot. There's no execution path at all.
- You're looking for signals to follow without reading the reasoning.
- You trade instruments outside US-listed equities and equity options.
The reasoning is the thing to judge.
Sign up and read one full recommendation on the sample book. It takes about a minute and costs nothing.
No credit card · 30-day trial with Solo features · Free tier stays free