> 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/how-pharos-does-it/resolution.md).

# Resolution

Discover how Pharos's AI layer — Org Genie and MCP-connected AI agents —  helps identify root causes and suggest remediations to accelerate issue  resolution.

Resolution is where the rubber meets the road — taking prioritized issues and actually fixing them. Pharos provides AI-powered analysis, plain-language AI-agent investigation, visual debugging tools, and tracing capabilities to dramatically reduce the time from "we know there's a problem" to "the fix is deployed."

***

## The Challenge: Time-Consuming Investigation

Even with perfect tracking and triage, developers still spend significant time on investigation: reproducing issues from vague error messages, gathering context about what the code was trying to do, identifying which recent changes might have introduced the failure, and understanding the business impact and urgency.

{% hint style="danger" %}
**Traditional MTTR:** Days to weeks. **With Pharos:** Minutes to hours.

The difference comes from giving developers the context, analysis, and tools they need from the start — instead of making them hunt for it.
{% endhint %}

***

## The Pharos Resolution Toolkit

{% tabs %}
{% tab title="🤖 AI Analysis" %}
**Instant insights for every Issue.** [Org Genie](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/org-genie.md) generates AI-powered analysis that gives your team a head start on investigation, proposing root cause and remediation at the issue level.

**Root Cause Hypothesis:**

* What likely caused this error?
* Which code paths or configuration are involved?
* Are there related metadata changes?

**Remediation Suggestions:**

* Specific code changes to consider
* Configuration adjustments to make
* Testing strategies to prevent recurrence

**Business Context:**

* Why this matters — user impact, data impact, and revenue risk
* Similar past incidents and how they were resolved
* Trends indicating whether the problem is growing

→ [Learn more about AI Analysis](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/issue-analysis.md)
{% endtab %}

{% tab title="💬 Org Genie" %}
**Conversational AI for complex investigations.** When an Issue requires deeper exploration, Org Genie lets you ask questions in natural language.

**Capabilities:**

* Ask questions about the Issue and get contextual answers
* Query org metadata and configuration
* Get explanations of error patterns and relationships
* Explore "what if" scenarios interactively

**Example Conversation:**

```
You: "Why is this DML exception only affecting these specific users?"

Org Genie: "The affected users share the 'Sales Rep' profile, which 
lacks Edit permission on the Industry field. The validation rule 
'Require_Industry' was modified 2 weeks ago to make this field 
required. This timing aligns with the Issue's first occurrence."

You: "Show me similar issues we've had in the past"

Org Genie: "There are 3 related issues:
- ISSUE-1234: Same validation rule, different object (resolved)
- ISSUE-2456: Permission issue on same profile (open)
- ISSUE-3789: Industry field requirement (closed as duplicate)"
```

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

{% tab title="🕵️ AI Agent (MCP)" %}
**Investigate by asking, in plain language.** Connect any MCP-capable [AI agent](https://docs.pharos.ai/configuration-guide/integrations/ai-agents-mcp.md) to Pharos and describe a symptom — no filters, SOQL, or knowledge of Pharos internals required.

**How it works:**

* You describe a symptom — *"why is checkout failing?"* or *"why did Issue #1234 start?"*
* The agent scopes the query from that symptom, correlates logs and recent metadata changes, and runs AI root-cause analysis — with no predefined criteria to match against
* It answers in plain language, and can walk the full loop — investigate, trace, triage, even open a ticket
* **Every write is proposed for your confirmation first** — the agent digs and recommends, a human approves the change

→ [Learn more about AI agents and MCP](https://docs.pharos.ai/configuration-guide/integrations/ai-agents-mcp.md)
{% endtab %}

{% tab title="🔍 Visual Debugging" %}
**Transform cryptic errors into actionable insights.** Pharos provides specialized debuggers for each error context.

**Flow Debugger:**

* Interactive flow diagram showing the exact execution path
* Variable values at each decision point
* Exact node where the failure occurred, with related record context

**Apex Stack Trace Enhancer:**

* Code snippets for each stack frame
* Line-by-line context around the failure point
* Variable state at the moment of failure

**Bulk API Analyzer:**

* Summary view with distinct errors and their counts
* Detail view showing every failed record with its specific error
* CSV export for batch remediation

→ [Learn more about Log Presentations](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/logs/log-presentations.md)
{% endtab %}

{% tab title="📡 Tracing" %}
**Real-time debugging for intermittent and hard-to-reproduce issues.**

**Issue Tracing:**

* Schedule debug log capture for specific users or scenarios
* Predictive tracing anticipates when the Issue will recur
* Automatic debug log retrieval and analysis
* Results linked to the Issue for full context

**Transaction Tracing (via Triton):**

* Follow execution across LWC → Apex → external systems
* Performance metrics at each step in the chain
* Request/response payload capture for integration troubleshooting

→ [Learn more about Tracing](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/tracing.md)
{% endtab %}
{% endtabs %}

***

## Resolution Workflow

A typical resolution follows a structured path from analysis to verification:

{% hint style="info" %}
**Step 1: Review AI Analysis** Read the root cause hypothesis, check the remediation suggestions, and determine whether this is a quick fix or requires deeper investigation.
{% endhint %}

{% hint style="info" %}
**Step 2: Examine Debug Context** Review the visual debugger (Flow, Apex, or Bulk API), check related logs for patterns, and look at recent org changes in the audit trail.
{% endhint %}

{% hint style="info" %}
**Step 3: Ask, in plain language (if needed)** For complex issues, ask Org Genie directly in Pharos — or describe the symptom to a connected MCP AI agent and let it correlate logs and changes on its own: "Why might this be happening?", "What changed recently that could cause this?"
{% endhint %}

{% hint style="info" %}
**Step 4: Schedule Tracing (if needed)** For intermittent issues: set up a trace schedule, wait for the Issue to recur with debug logs captured, then analyze the trace results with full execution context.
{% endhint %}

{% hint style="info" %}
**Step 5: Implement and Deploy** Make code or configuration changes, test in sandbox, and deploy to production.
{% endhint %}

{% hint style="info" %}
**Step 6: Monitor for Recurrence** Set the Issue status to "Completed." Pharos monitors for new occurrences — if the error recurs, the Issue is automatically reopened and a new Jira ticket is created (if integration is enabled).
{% endhint %}

***

## Measuring Resolution Effectiveness

Track these metrics to improve your resolution process over time:

| Metric                      | Target                | Action If Not Meeting                             |
| --------------------------- | --------------------- | ------------------------------------------------- |
| **Mean Time to Resolution** | <24 hours (Critical)  | Improve AI analysis coverage, add tracing         |
| **First-Time Fix Rate**     | >85%                  | Better root cause analysis before deploying fixes |
| **Recurrence Rate**         | <15%                  | Improve testing, add regression tests             |
| **Time Saved per Issue**    | >2 hours vs. baseline | More team training on Pharos tools                |

***

## Closing the Loop

Resolution isn't the end — it's part of the continuous improvement loop. Each resolved Issue makes the team more effective at resolving the next one:

1. **Fix deployed** → Pharos monitors for recurrence
2. **Metrics improve** → system health trends upward
3. **Learnings captured** → team knowledge grows with every investigation
4. **Process refined** → the next issue gets resolved faster

{% hint style="success" %}
**The Complete Loop:** With Tracking, Awareness, Action, and Resolution working together, your team has transformed from reactive firefighting to proactive system stewardship.

**Next:** Explore [Pharos Platform Architecture](https://docs.pharos.ai/platform-overview/pharos-platform-architecture.md) for deep dives into each component.
{% endhint %}
