AI Insights

Basic & Pro Plans AI Assistant Required

The AI Insights feature uses artificial intelligence to analyze your alerts vs fills comparison data and provide actionable recommendations for improving execution quality.

---

Overview

AI Insights examines your execution data and provides:

• Pattern recognition — Identifies trends and anomalies
• Root cause analysis — Explains why issues occur
• Optimization recommendations — Actionable steps to improve
• Strategy suggestions — Code-level improvements

---

Prerequisites

Before using AI Insights, ensure:

• Basic or Pro plan — Required for AI features
• AI Assistant configured — Set up in Account Settings
• Comparison data available — Generate an alerts vs fills comparison first
• Sufficient credits — AI insights consume credits based on analysis size

Caution

No comparison data? AI Insights requires completed comparison data. Generate your alerts vs fills analysis first, then click “Analyze with AI”.

---

How to Use AI Insights

1. Generate Comparison Data

   Complete your alerts vs fills analysis:

   • Select symbol and date range
   • Load TradingView alerts (webhook or CSV)
   • Load Alpaca fills
   • Wait for matching to complete

2. Open AI Insights

   Click the “Analyze with AI” button in the header:

   • Button appears after comparison is complete
   • Shows loading state while generating
   • Opens the AI Insights drawer

3. Review Analysis

   The AI Insights drawer displays:

   • Executive summary — High-level findings
   • Key insights — Top issues identified
   • Recommendations — Actionable steps
   • Reasoning trail — How AI reached conclusions

4. Take Action

   Use the insights to:

   • Copy prompt — Use in Pine Script AI Editor
   • Open in Editor — Send context to AI editor
   • Save snapshot — Preserve analysis with insights

---

Understanding AI Insights

Executive Summary

The summary provides:

• Overall assessment — Good, acceptable, or concerning
• Primary issues — Top 2-3 problems identified
• Impact estimate — How much these issues cost you

Key Insights Section

Each insight includes:

• Issue description — What the problem is
• Evidence — Data supporting the finding
• Frequency — How often it occurs
• Impact — Financial or performance cost

Recommendations

AI provides specific, actionable recommendations:

Category	Examples
Infrastructure	Network optimization, server location
Order Types	Switch to limit orders, adjust timing
Strategy	Code changes, parameter adjustments
Broker	API settings, account configuration

Reasoning Trail

The reasoning trail shows:

• Step-by-step analysis — How AI processed your data
• Data points considered — What metrics were analyzed
• Conclusion logic — Why recommendations were made

Tip

Review the reasoning trail to understand how AI reached its conclusions. This helps you validate the recommendations and learn from the analysis.

---

Opening in Pine Script AI Editor

You can send your analysis context directly to the Pine Script AI Editor for code-level optimization.

How It Works

1. Click “Open in Pine Script AI” in the AI Insights drawer
2. Context is prepared — Analysis data is formatted for AI
3. Editor opens — Pine Script AI Editor loads with context
4. Chat is pre-populated — Initial prompt includes your analysis

What Gets Sent

The context includes:

• Comparison summary — Match rate, latency, slippage
• Key metrics — Top issues and patterns
• Trade examples — Sample problematic trades
• AI insights — Generated recommendations

Use Cases

1. Code optimization — Get specific Pine Script improvements
2. Strategy refinement — Adjust strategy based on execution data
3. Parameter tuning — Optimize strategy settings
4. Bug fixes — Identify and fix code issues

---

AI Insights Workflow

Standard Workflow

Generate Comparison → Analyze with AI → Review Insights →
Take Action → Monitor Results → Repeat

Advanced Workflow

1. Generate comparison for a time period
2. Save snapshot with baseline metrics
3. Generate AI insights and recommendations
4. Implement changes based on recommendations
5. Generate new comparison after changes
6. Compare snapshots to measure improvement
7. Iterate until execution quality improves

---

Best Practices

When to Use AI Insights

Good times:

• After completing a comparison with concerning metrics
• When you see patterns but don’t understand them
• Before making significant strategy changes
• When comparing different time periods

Less useful:

• With very small datasets (<10 trades)
• When metrics are already excellent
• Without comparison data

Getting Better Insights

1. Larger datasets — More trades = better pattern recognition
2. Complete data — Ensure all trades are loaded
3. Relevant timeframes — Use recent, representative periods
4. Clear questions — Know what you want to optimize

Interpreting Results

• Consider context — AI doesn’t know your full strategy
• Validate recommendations — Test changes before deploying
• Review reasoning — Understand why AI suggests changes
• Combine with expertise — Use AI as a tool, not replacement

---

Limitations

What AI Can’t Do

• Predict future performance — Only analyzes past data
• Account for all factors — May miss external influences
• Guarantee improvements — Recommendations are suggestions
• Replace human judgment — Use as input, not final decision

Data Quality Matters

AI insights are only as good as your data:

• Incomplete data → Incomplete insights
• Inaccurate data → Misleading recommendations
• Small datasets → Less reliable patterns
• Outdated data → Irrelevant suggestions

---

Troubleshooting

”AI Insights Unavailable”

Causes:

• No comparison data generated
• AI Assistant not configured
• Insufficient credits
• Service temporarily unavailable

Solutions:

1. Generate comparison data first
2. Configure AI Assistant in Account Settings
3. Purchase credits if needed
4. Try again later

”Loading AI Credentials”

Meaning: System is checking your AI configuration.

Wait time: Usually 1-2 seconds.

If it persists: Check Account Settings → AI Assistant configuration.

Insights Seem Incorrect

Possible reasons:

• Small dataset (not enough data)
• Unusual market conditions
• Data quality issues

Actions:

• Review reasoning trail
• Check data completeness
• Generate insights for different time period
• Validate with manual analysis

---

Related Sections

• Advanced Analytics — Detailed analysis tools
• Metrics & Analysis — Understanding metrics
• Pine Script Editor — Code optimization
• Account Settings — AI Assistant configuration