Real-time price tracking, AI predictions, and data quality explained.
The Live Market Dashboard lets you track a custom list of stocks with live prices refreshed every 15 seconds. It shows price, dollar/percent change, an AI-generated directional prediction, a data quality score, and the time of the last update — all in a single table.
Data is fetched via Yahoo Finance (proxied through the TradePro Cloudflare Worker) — no API key required. If you have FMP or TwelveData keys configured, they can serve as fallbacks.
TSLA) in the Add Symbol field and press Enter or click Add.Symbol — The stock ticker you added (e.g. AAPL, TSLA).
Price — Current market price in USD, as reported by Yahoo Finance.
How much the stock has moved compared to yesterday's closing price.
Previous close is fetched from Yahoo Finance's chartPreviousClose field, which matches what you see on finance.yahoo.com.
A rules-based directional signal generated from recent price ticks collected since you added the symbol.
| Label | Meaning |
|---|---|
| 📈 Bullish | Recent trend and momentum suggest upward price movement |
| 📉 Bearish | Recent trend and momentum suggest downward price movement |
| ➡️ Neutral | No clear directional signal in recent ticks |
| Analyzing... | Fewer than 10 data points collected — not enough to generate a signal |
Important: The AI only looks at price ticks since you added the symbol in this session. It has no access to historical OHLCV data. It reflects short-term intraday momentum only.
How confident the AI is in its prediction (0–100%). Starts low and builds as more ticks are collected.
The model calibrates itself using a 65% success threshold — predictions that historically resolve correctly ≥65% of the time are considered reliable enough to show.
A colored dot indicating how reliable the current data point is, based on API response time, data completeness, and data age.
| Dot | Score | Meaning |
|---|---|---|
| Green | ≥80% | Fresh, complete data — high reliability |
| Yellow | 60–79% | Acceptable quality — minor latency or partial data |
| Red | <60% | Low quality — stale, slow, or incomplete response |
Timestamp (HH:MM:SS) when the last data point was received. Normally refreshes every 15 seconds during market hours.
The AI uses a combination of linear regression on price history and momentum scoring. Here are the exact thresholds:
| Condition | Signal | Initial Confidence |
|---|---|---|
| Change % > +0.5% | 📈 Bullish | ~40–60% |
| Change % < -0.5% | 📉 Bearish | ~40–60% |
| |Change %| < 0.5% | ➡️ Neutral | ~40% |
| |Change %| > 2% | Extra pattern scan triggered | Varies |
| Linear regression slope > 0.001 | Trend pattern detected (bullish/bearish) | Varies |
| Trend direction flips mid-window | Reversal pattern detected | Varies |
When enough data is available, the AI combines multiple signals using these weights:
| Factor | Weight |
|---|---|
| Price momentum | 25% |
| Momentum score | 25% |
| Volume patterns | 20% |
| Volatility | 15% |
| Technical indicators | 15% |
Data is fetched in this priority order:
When 10+ ticks are available, the AI scans for four pattern types:
| Pattern | What It Means |
|---|---|
| Trend | Consistent upward or downward price slope across recent ticks (linear regression) |
| Reversal | Trend direction switched in the latter half of the observation window |
| Breakout | Price moves significantly beyond recent range |
| Support/Resistance | Price repeatedly bouncing near the same level |