ML Alpha

The Screener covers over 5,000 US publicly traded companies with more than 25 years of historical data. You filter on fundamental and technical factors, and each result carries a description, financial charts and statements. Machine learning predictions sit alongside the raw numbers, so a screen can mix classic value or momentum rules with the model's view.

AI Top Picks turns the platform's ML Alpha Scores into a shortlist tied to your stated investment goals. Instead of starting from a blank screener, you get a set of stocks the model rates highest for your objectives. The scores are the headline number here, and they come with verified past performance rather than a black-box claim.

The AI Financial Expert is a generative AI helper that explains a company's financial state in plain language. You ask about a stock and it walks through the numbers, so you can check the health of a company without reading every line of a 10-K. It pairs with the screener, which means the analysis points back to the data you were already looking at.

The Data Science Studio lets advanced users build their own machine learning models through a point-and-click interface, with no coding required. You pick the data and the prediction target, then the platform handles the training. It ships with the curated ML-ready datasets ML Alpha uses for its own models, so you start from the same material.

You can backtest any screening approach against historical data to see how it would have behaved. The platform compares a strategy's performance against the market benchmark, which answers the question most stock tools sidestep: did this rule actually beat the index, or just feel clever? The free plan includes 15 backtesting credits per month, and paid tiers raise that limit.

The Marketplace is where the community layer lives. Investors and data scientists list portfolios, and subscribers pay to follow them live, with notifications whenever the owner changes a holding. Sellers set their own price, and the platform's example math lands at $10 per month per follower. A Portfolio Leaderboard ranks performance so you can see who is worth following.

The meta model scores stocks on five criteria: short, mid and long-term performance, dividend growth, and volatility. You adjust how much weight each one carries, so a dividend hunter and a momentum trader can dial in very different models from the same engine. It is the part of the platform that rewards people who know what they want from the market.