About
One person,
and a rule.
I’m Giacomo Frassetto. I build Bot Marketplace on my own, from Italy. Not a fund, not a trading desk, not a team of twelve in a stock photograph.
Why it exists
Because you cannot watch charts all day, and things happen while you’re away. That’s the entire problem I set out to solve: tools that watch for you, tell you when something happens, and then get out of the way. What you do with the alert is your call — and it always was going to be, because I don’t know your situation and nobody selling you certainty does either.
The rule
Nothing on this site
that I can’t show you.
It sounds like the sort of thing everybody writes, so here is what it actually cost me.
I deleted the machine learning
The site used to promise that Premium bots trained a model on each pair’s history, and printed things like “84% confidence”. There was no model. There was a twelve-line function reading how far past a threshold an indicator had gone. What replaced it is a signal strength rating that says only what it can: how this alert compares with the previous ones of the same strategy on the same pair — “sharper than 80% of them”. That is a ranking, not a chance of success, and the alert says so.
I published the failures
9 strategies, 83 measured combinations of pair and timeframe, and 6 that beat random timing. The other 77 are on the Signals page with their real numbers. Hiding them would have made the rest worth nothing.
I threw away a 67.6%
One strategy measured 67.6% accuracy. True, and measured — on 37 alerts, where the uncertainty is about ten points either way and a real edge here is three to five. So now nothing gets a published percentage below 100 alerts, and that number never went on the site.
I don’t track whether you read this
No pixel in our emails telling me you opened them, no link that passes through a counter before going where it says. A product whose whole promise is “check it yourself” has no business watching who checks.
I can’t compete with the people selling 90% win rates, and I stopped trying. What I can do is be the one place where the numbers come with their margin of error and the link to check them. That’s a smaller promise. It’s also the only one I can keep.
Where I deliberately stop
These aren’t features waiting to be built. They are decisions.
No orders, ever
I never ask for exchange API keys and the software cannot place a trade. Holding keys to other people’s money means being a target and, in the EU, a regulated activity. Both are reasons enough.
No advice
No recommendations, no price targets, no portfolio management. Everything I send is identical for every subscriber with the same settings, by design.
No predictions
I report what already happened — a condition met, a document filed, an amount moved. Nobody knows the rest.
Ask me anything about it.
The same address answers whether you’re a customer or just curious, and it’s me reading it.
We don’t predict the future. We read the present.