Find Stock Opportunities with AI

What the Opportunity Radar sweeps

One scan covers 50 assets: 18 crypto pairs, 16 US stocks and 16 Borsa Istanbul shares. The sweep runs on the daily timeframe, which is long enough to filter out intraday noise and short enough to surface setups while they are still forming.

The point is coverage without effort. Opening fifty charts by hand to find the three worth studying is an hour you spend before you have learned anything; the scan compresses that into a ranked list.

Stage one runs with no AI at all

The first stage is pure arithmetic on the server. Candles are fetched for each symbol, the indicator set is computed, and the score is calculated locally — not one model call is made during this pass, which is why a full sweep takes roughly three seconds.

That architecture is also why a scan can be offered on the free plan at all. The expensive part of AI analysis is the model, and the screening stage simply does not use one.

The 0-100 score and its five sub-scores

Every asset receives an opportunity score between 0 and 100, assembled from five components. Trend reads the moving average arrangement, the position of price relative to its long-term average and the strength and direction of ADX. Momentum measures the distance of RSI from the midpoint, pulling its own contribution back at the extremes because an overstretched reading is a risk rather than an opportunity.

Volume compares current activity against its recent average as confirmation. Technical position asks where price sits relative to its bands and to the nearest support and resistance zones, which is what determines whether an entry has room. Risk is scored so that a high number means low risk, penalising wide volatility and extreme oscillator readings.

Trend and momentum carry the most weight, volume and position confirm, and risk applies a smaller correction. A high score therefore means a technically pronounced situation, not a recommendation.

Stage two: six candidates, one request

Only the six highest-scoring assets go to the model, and they go together in a single request rather than one call each. Each comes back with a short comment explaining what the numbers add up to.

The cards below the top six are still listed with their full scores and sub-scores. They simply carry the local reading without an AI comment attached, which keeps the scan fast and its cost predictable.

Reading and filtering the results

Up to twelve cards are shown. Each one displays the current price, the opportunity score, the trend, momentum and risk labels, and the specific reasons the setup was flagged — a moving average alignment, an ADX reading indicating a strong trend, volume clearly above its twenty-bar average, and so on.

Two filters narrow the list: market and risk level. Filtering to a single market is the fastest way to compare like with like, since the volatility profile of a crypto pair and of a large-cap share are not on the same scale.

A high score is not a direction

The score measures how pronounced a technical situation is, and a decisive downtrend is as pronounced as a decisive uptrend. That is why every card carries its trend and momentum labels separately rather than collapsing into one verdict; the score tells you where to look, the sub-scores tell you what you are looking at.

From the screener into a full analysis

The scan is the shortlist, not the conclusion. Once a card has earned your attention, run the full analysis on that symbol to get the entry, stop and target levels, the risk/reward ratio and the invalidation condition, on whichever timeframe matches your horizon.

The free plan includes one scan per day and Pro raises it to twenty, which is the difference between checking the market once each morning and re-checking it as conditions change.