> For the complete documentation index, see [llms.txt](https://docs.pharos.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis.md).

# AI Analysis

How Pharos AI analysis combines observability data with Salesforce metadata to deliver intelligent, context-aware troubleshooting insights.

Pharos integrates AI-powered analysis throughout the platform to accelerate troubleshooting and provide deep insights into your Salesforce environment. Unlike generic AI tools, Pharos AI—powered by **Org Genie**—works exclusively with trusted data from your Salesforce org, ensuring accuracy and relevance without relying on external training data.

{% hint style="success" %}
**From Days to Minutes**

AI analysis transforms error investigation from a manual, time-consuming process into an intelligent, guided workflow. What once took hours of report building and log analysis now happens in seconds with natural language queries.
{% endhint %}

***

## Why Pharos AI is Different

Traditional observability platforms require you to know exactly what you're looking for before you can find it. You build dashboards, create custom reports, filter list views—all while manually correlating errors with recent changes.

Pharos AI inverts this paradigm by enabling **conversational observability**—ask questions in plain language and receive intelligent, context-aware answers grounded in your actual org data.

### Key Differentiators

The differences show up across the entire troubleshooting workflow, from where data comes from to how quickly you reach an insight.

| Feature             | Traditional Approach       | Pharos AI                                                     |
| ------------------- | -------------------------- | ------------------------------------------------------------- |
| **Data Source**     | Generic external models    | Your Salesforce org data exclusively                          |
| **Understanding**   | Basic error interpretation | Salesforce-native concepts (governor limits, triggers, flows) |
| **Correlation**     | Manual cross-referencing   | Automatic change correlation with error patterns              |
| **Interface**       | Dashboard building         | Natural language conversation                                 |
| **Time to Insight** | Hours of investigation     | Seconds to minutes                                            |

{% hint style="info" %}
**Your Data, Your Org, Your Answers**

Org Genie analyzes data entirely within your Salesforce org—logs, metadata, configuration, and audit trails. It doesn't rely on external data sources or generic training models. The AI understands **your specific environment**, including custom objects, fields, automation, and business logic.
{% endhint %}

***

## AI Analysis Capabilities

Pharos provides two complementary AI capabilities, each designed for specific use cases:

### 1. Issue Analysis: Automated Root Cause Investigation

Every Issue in Pharos can be automatically analyzed to provide structured insights that accelerate troubleshooting.

**What You Get:**

* 🎯 **Severity Assessment** - AI-driven prioritization based on impact, frequency, and affected data
* 💡 **Problem Explanation** - Clear descriptions interpreting technical errors in accessible language
* 🔍 **Root Cause Indicators** - Correlated metadata changes and related error patterns
* ⚠️ **Impact Analysis** - Business implications and cascading effects
* ✅ **Recommended Actions** - Practical, context-specific next steps

**Automation Options:**

* Deliver analysis automatically to Slack, Teams, or Jira when Issues meet specific criteria
* Trigger analysis via Salesforce Flows for custom triaging workflows
* Schedule analysis for regular Issue reviews and health checks

**Example Use Case:**\
A production error spike occurs after deployment. Instead of spending 2-3 hours manually reviewing logs and audit trails, AI analysis identifies the problematic validation rule change in under 5 minutes, showing exactly when it was deployed and how it correlates with the error spike.

→ [**Learn more about Issue Analysis**](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/issue-analysis.md)

### 2. Org Genie: Conversational Observability

Ask natural language questions about your entire Salesforce environment and receive intelligent answers with visualizations—no dashboard building required.

**Example Queries:**

```
"Show me all Apex exceptions related to OpportunityTrigger in the last 7 days"
"What configuration changes were made last week?"
"Which Issues from the last 30 days warrant immediate attention?"
"Is there a correlation between yesterday's deployment and today's error spike?"
```

**Core Capabilities:**

* 📊 **Generate Visualizations** - Create charts on demand to spot trends and anomalies
* 🔗 **Correlate Changes** - Connect metadata modifications with error patterns automatically
* 🔍 **Query Across Data** - Search logs, Issues, and metadata simultaneously
* 📈 **Identify Trends** - Spot emerging patterns before they become critical
* 🎯 **Navigate Intelligently** - Start broad, then drill into specific problems

**Two Contexts for Different Needs:**

{% tabs %}
{% tab title="🌐 Org-Level Queries" %}
**Access from:** Main navigation menu

**Best for:**

* Analyzing trends across your entire environment
* Comparing time periods ("this week vs last week")
* Prioritizing which Issues deserve attention
* Weekly health checks and sprint planning

**Example:** *"Which Issues from the last 7 days warrant immediate attention and why?"*

→ [**Learn more about Org-Level Queries**](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/org-genie/org-level.md)
{% endtab %}

{% tab title="🔍 Issue-Level Queries" %}
**Access from:** Any Issue detail page

**Best for:**

* Investigating why a specific Issue started or stopped
* Correlating metadata changes with a particular error pattern
* Understanding if a fix actually resolved the problem
* Deep-dive root cause analysis

**Example:** *"This Issue stopped occurring on December 17th. Were there any metadata changes that could have fixed it?"*

→ [**Learn more about Issue-Level Queries**](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/org-genie/issue-level.md)
{% endtab %}
{% endtabs %}

→ [**Learn more about Org Genie**](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/org-genie.md)

***

## Getting Started

Teams get the most out of AI analysis by adopting it progressively — reading structured analysis on issues you already have, then asking Org Genie exploratory questions, then automating delivery once you trust the output. [Getting Started with AI Analysis](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/issue-analysis.md#getting-started-with-ai-analysis) lays out that path in stages.

To set the features up, work through the checklist in [Org Genie Settings](https://docs.pharos.ai/configuration-guide/org-genie-settings.md); OpenAI is the one hard requirement, and enabling [Change Tracking](https://docs.pharos.ai/configuration-guide/change-tracking.md) is what unlocks correlation between errors and configuration changes.

***

## Data Privacy and Security

Pharos AI operates with strict data boundaries: it analyzes error metadata, configuration structure, and statistical aggregations, and never sends your actual record data, transaction details, or restricted fields. Your data is not used to train external models, the index is stored in vectorized form, and access follows your existing Salesforce sharing model.

For the reasoning behind that boundary see [The Pharos Trust Layer](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/issue-analysis.md#the-pharos-trust-layer); for the itemized list of what is and isn't sent, see [Security & Data Access](https://docs.pharos.ai/configuration-guide/security.md#ai-data-handling).

***

## Automation and Integration

While on-demand AI analysis is powerful, the real transformation happens when you integrate it into your team's workflow:

**Automated Delivery:**

* 💬 **Slack** - Threaded messages with analysis to designated channels
* 📧 **Microsoft Teams** - Adaptive cards keeping ops and dev teams informed
* 🎫 **Jira** - Automatic ticket comments enriching context

**Automation Triggers:**

* **Threshold-Based** - Log count exceeds limits, user impact reaches critical levels
* **Status Changes** - New Issues, resolved Issues regressing, severity escalations
* **Pattern Detection** - Sudden error spikes, new error types, anomalies

**Flow Integration:**

* Invoke AI analysis from Salesforce Flows for custom triaging workflows
* Build smart escalation processes based on AI severity assessment
* Implement post-deployment verification flows

→ [**Learn more about Issue Analysis Automation**](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/issue-analysis.md#automating-ai-analysis)\
→ [**Learn more about Alerting Configuration**](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/alerting.md)

***

## Key Considerations

A few things are worth keeping in mind to get reliable results from AI analysis.

{% hint style="warning" %}
**AI is Probabilistic, Not Deterministic**

Org Genie provides **likely** root causes and **suggested** next steps based on patterns. Always validate AI suggestions with manual review before taking action, especially for critical systems.
{% endhint %}

**For Best Results:**

* Use [Triton custom instrumentation](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/logs/log-ingestion-methods/pharos-triton-instrumentation.md) to log rich contextual data
* Enable Change Tracking for metadata correlation capabilities
* Combine AI insights with your domain knowledge and technical expertise
* Allow time for historical data to accumulate for more accurate trend analysis

{% hint style="info" %}
**AI as a Tool, Not a Replacement**

AI analysis accelerates your troubleshooting and surfaces insights you might miss. Final decisions about prioritization and resolution remain with your team, informed by AI-generated intelligence but guided by your business knowledge and technical expertise.
{% endhint %}

***

## In This Section

Dive deeper into specific AI analysis capabilities:

<table data-view="cards"><thead><tr><th></th><th></th><th data-hidden data-card-target data-type="content-ref"></th></tr></thead><tbody><tr><td><strong>🔍 Issue Analysis</strong></td><td>Automated AI analysis for individual Issues—severity, root cause, and recommendations.</td><td><a href="https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/issue-analysis.md">Issue Analysis</a></td></tr><tr><td><strong>🧞 Org Genie</strong></td><td>Conversational AI queries across your org, from health checks to deep troubleshooting.</td><td><a href="https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/org-genie.md">Org Genie</a></td></tr></tbody></table>

***

{% hint style="success" %}
**Start Experimenting Today**

1. Open any Issue in Pharos
2. Navigate to the "AI Analysis" tab
3. Review the auto-generated insights
4. Click into Org Genie and ask your first question
5. Request a visualization to see patterns emerge

The best way to understand AI analysis is to use it. Start with simple questions and build from there.
{% endhint %}
