> 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/pharos-philosophy.md).

# Pharos Philosophy

Understand the tracking → awareness → action and collaboration → resolution loop  that forms the foundation of Pharos's approach to observability.

Pharos doesn't just collect errors — it guides your team through a structured loop that turns raw operational signals into resolved issues and lasting improvements. And the loop isn't limited to errors: application exceptions, governor-limit pressure, Event Monitoring signals, metadata changes, and custom business events all flow through the same four stages — **Tracking**, **Awareness**, **Action & Collaboration**, and **Resolution** — before feeding back into the cycle. **AI runs through every stage, and the loop stays open at every step** — the two themes that make this more than a linear error workflow.

Each stage narrows raw signal into resolved outcome, backed by the tools your team already uses:

![The Pharos loop — Tracking (Issues, Logs), Awareness (Slack, Teams, Email), Action and Collaboration (Jira, Azure DevOps), and Resolution (AI-powered issue analysis)](https://4264222071-files.gitbook.io/~/files/v0/b/gitbook-x-prod.appspot.com/o/spaces%2Fdl9tmJ56lvwg5XY63QYr%2Fuploads%2Fgit-blob-9679605e740c50e6ac265da8897839cbc9dfbf2b%2Fpharos-loop-funnel.png?alt=media)

***

## The Four Stages

{% tabs %}
{% tab title="📡 Tracking" %}
**Goal:** Capture every failure with full context — automatically.

Pharos begins by collecting errors and events across your Salesforce environment without requiring manual instrumentation:

* **Automated error collection** — zero-config ingestion captures platform errors out of the box
* **Custom instrumentation** — extend tracking to cover business-critical processes and custom logic
* **External system integration** — pull in signals from connected systems for a complete picture

{% hint style="info" %}
**Output:** Enriched log records with contextual metadata — who was affected, what they were doing, and what went wrong.
{% endhint %}
{% endtab %}

{% tab title="🔔 Awareness" %}
**Goal:** Alert the right people at the right time — without the noise.

Raw errors are worthless if they drown your team in alerts. Pharos transforms log data into targeted, meaningful notifications:

* **Intelligent grouping** — deduplicates related errors so teams aren't bombarded with repeats
* **Context-rich notifications** — every alert includes the information needed to understand severity and impact at a glance
* **Multi-channel delivery** — routes alerts through Slack, Teams, PagerDuty, or email based on team preferences

{% hint style="info" %}
**Output:** Actionable alerts routed to the appropriate teams — not just noise, but signals that demand attention.
{% endhint %}
{% endtab %}

{% tab title="🎬 Action & Collaboration" %}
**Goal:** Triage and prioritize systematically, not reactively.

Once the right people are aware, Pharos provides the tools to decide what to fix first and coordinate the work:

* **Data-driven impact analysis** — understand which errors affect the most users, the most critical processes, or the highest-value workflows
* **Development workflow integration** — create and link issues in Jira or Azure DevOps directly from Pharos
* **Collaborative resolution tools** — share context, assign ownership, and track progress without leaving the platform

{% hint style="info" %}
**Output:** Prioritized issues with clear ownership — every team member knows what to work on and why.
{% endhint %}
{% endtab %}

{% tab title="✅ Resolution" %}
**Goal:** Fix root causes efficiently and verify the fix holds.

The final stage equips your team with the tools to resolve issues quickly and confirm they stay resolved:

* **AI-powered analysis** — [Org Genie](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/org-genie.md) surfaces root-cause recommendations and suggested fixes at the issue level
* **Plain-language investigation** — a connected [AI agent via MCP](https://docs.pharos.ai/configuration-guide/integrations/ai-agents-mcp.md) can be asked a symptom in natural language and will correlate logs and changes itself
* **Debug tools** — trace execution paths, inspect variable states, and use visual debuggers to pinpoint problems
* **Verification and monitoring** — confirm the fix is deployed and watch for recurrence automatically

{% hint style="info" %}
**Output:** Deployed fixes with recurrence monitoring — the issue is resolved, and Pharos keeps watching.
{% endhint %}
{% endtab %}
{% endtabs %}

***

## Why This Matters

Traditional error management is reactive: wait for user complaints, scramble to investigate, and hope you find the root cause before the next escalation. The Pharos loop flips this model — detect issues early, prioritize based on real impact, and resolve them systematically.

{% hint style="success" %}
**The Result:**

* **Reduced MTTR** — faster detection and richer context shorten time-to-resolution
* **Fewer production incidents** — proactive monitoring catches problems before users notice
* **Better resource allocation** — impact-based prioritization ensures engineering effort goes where it matters most
* **Improving reliability over time** — each cycle feeds lessons back into your observability practice
  {% endhint %}

***

## Intelligent and Open at Every Stage

Two things keep the loop from being a rigid, manual pipeline.

**It's intelligent.** [AI](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis.md) runs across the whole cycle rather than sitting at the end — [Org Genie](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/ai-analysis/org-genie.md) proposes root cause and remediation at the issue level and surfaces patterns at the org level, and a connected [AI agent via MCP](https://docs.pharos.ai/configuration-guide/integrations/ai-agents-mcp.md) can walk the entire loop conversationally: investigate, trace, and triage in plain language, with every write proposed for human confirmation first. AI shortens the path from Tracking to Resolution without removing the human from the decision. See [Intelligence](https://docs.pharos.ai/platform-overview/how-pharos-does-it/intelligence.md) for how this works stage by stage.

**It's open.** The loop isn't locked inside Pharos. Telemetry captured in Tracking can flow out as [OpenTelemetry](https://docs.pharos.ai/open-telemetry/open-telemetry.md) to the observability backends your teams already run *(early-access)*, and the platform itself is built to be operated by AI agents and extended through native APIs, platform events, webhooks, and a rules engine — see [Extensibility & Integrations](https://docs.pharos.ai/platform-overview/pharos-platform-architecture/extensibility.md). Native and closed are not the same thing: Pharos stays open at every stage of the cycle.

***

## Continuous Improvement

The loop doesn't end at resolution. Once a fix is deployed, Pharos continues monitoring for recurrence and feeding trend data back into the tracking layer. Over time, this creates a compounding effect — each cycle sharpens your team's ability to detect, prioritize, and resolve issues faster than the last.

{% hint style="info" %}
**Next Steps:** Explore each stage in detail to understand how Pharos's components work together to power this loop.
{% endhint %}
