| 1 | --- |
| 2 | title: Roles & mental model |
| 3 | description: How to think about CoPilot as a SOC operator vs. admin/engineer vs. developer. |
| 4 | --- |
| 5 | |
| 6 | CoPilot is easiest to learn if you separate it into **two jobs**: |
| 7 | |
| 8 | 1) **Operate incidents** (alerts → cases → evidence → response) |
| 9 | 2) **Make incidents possible** (connect sources/integrations so alerts flow) |
| 10 | |
| 11 | --- |
| 12 | |
| 13 | ## The mental model (visual) |
| 14 | |
| 15 | CoPilot becomes intuitive when you see it as **two loops** that share the same data: |
| 16 | |
| 17 | <Columns cols={2}> |
| 18 | <Card title="Admin / Platform loop" icon="gear"> |
| 19 | You make detection possible. |
| 20 | |
| 21 | <Steps> |
| 22 | <Step title="Connect & verify"> |
| 23 | Integrations + syslog connectors are healthy. |
| 24 | </Step> |
| 25 | <Step title="Ingest & route"> |
| 26 | Streams/pipelines route logs where they should go. |
| 27 | </Step> |
| 28 | <Step title="Provision & visualize"> |
| 29 | Customers/tenants, dashboards, indices, retention. |
| 30 | </Step> |
| 31 | <Step title="Enable alerting"> |
| 32 | Event definitions + notifications are configured. |
| 33 | </Step> |
| 34 | <Step title="Tune & maintain"> |
| 35 | Reduce noise, validate coverage, keep things reliable. |
| 36 | </Step> |
| 37 | </Steps> |
| 38 | </Card> |
| 39 | |
| 40 | <Card title="Operator loop" icon="siren"> |
| 41 | You run incidents. |
| 42 | |
| 43 | <Steps> |
| 44 | <Step title="Alert appears"> |
| 45 | New alert lands in Incident Management. |
| 46 | </Step> |
| 47 | <Step title="Triage"> |
| 48 | Decide: true positive? priority? scope? |
| 49 | </Step> |
| 50 | <Step title="Case work"> |
| 51 | Create a case, collect artifacts/evidence. |
| 52 | </Step> |
| 53 | <Step title="Respond"> |
| 54 | Contain, eradicate, recover. |
| 55 | </Step> |
| 56 | <Step title="Feedback"> |
| 57 | Feed improvements back into tuning/detections. |
| 58 | </Step> |
| 59 | </Steps> |
| 60 | </Card> |
| 61 | </Columns> |
| 62 | |
| 63 | ### How alerts become cases (the shared pipeline) |
| 64 | |
| 65 | <Steps> |
| 66 | <Step title="1) Data arrives"> |
| 67 | Endpoints (Wazuh), API integrations (O365/Mimecast/Huntress/CrowdStrike), and syslog devices (FortiGate/PAN‑OS/ASA). |
| 68 | </Step> |
| 69 | <Step title="2) Normalize & route"> |
| 70 | Graylog streams/pipelines normalize fields and route logs. |
| 71 | </Step> |
| 72 | <Step title="3) Detect"> |
| 73 | Graylog Event Definitions evaluate conditions and generate events. |
| 74 | </Step> |
| 75 | <Step title="4) Persist"> |
| 76 | Alerts are written to `gl-events*`. |
| 77 | </Step> |
| 78 | <Step title="5) Operate"> |
| 79 | CoPilot surfaces alerts → operators create/work cases → response. |
| 80 | </Step> |
| 81 | </Steps> |
| 82 | |
| 83 | ### “Where do I click?” (quick map) |
| 84 | |
| 85 | <Columns cols={3}> |
| 86 | <Card title="Operators" icon="siren" href="/user/operators-quickstart"> |
| 87 | Incident Management → Alerts → Cases |
| 88 | </Card> |
| 89 | <Card title="Admins / engineers" icon="gear" href="/user/admins-quickstart"> |
| 90 | Provisioning → Integrations → Connectors → Indices |
| 91 | </Card> |
| 92 | <Card title="Developers" icon="code" href="/developer/start-here"> |
| 93 | Architecture → Data flows → Connectors |
| 94 | </Card> |
| 95 | </Columns> |
| 96 | |
| 97 | --- |
| 98 | |
| 99 | ## SOC operator / analyst |
| 100 | |
| 101 | You spend most of your time in **Incident Management**. |
| 102 | |
| 103 | You care about: |
| 104 | - Is this alert real? |
| 105 | - What’s the blast radius? |
| 106 | - What evidence do I need? |
| 107 | - What do I do next? |
| 108 | |
| 109 | **Start here:** |
| 110 | - [Operator quickstart](/user/operators-quickstart) |
| 111 | - [Incident alerts (UI reference)](/user/ui/incident-alerts) |
| 112 | - [Cases (UI reference)](/user/ui/incident-cases) |
| 113 | - Videos: [Operator track](/user/videos#operator-track) |
| 114 | |
| 115 | ## Admin / engineer |
| 116 | |
| 117 | You care about **data onboarding and reliability**. |
| 118 | |
| 119 | You care about: |
| 120 | - Are sources connected and healthy? |
| 121 | - Are parsing / streams / pipelines configured correctly? |
| 122 | - Are indices and retention behaving? |
| 123 | - Is each customer/tenant provisioned correctly? |
| 124 | |
| 125 | **Start here:** |
| 126 | - [Admin/Engineer quickstart](/user/admins-quickstart) |
| 127 | - [Customer provisioning](/user/customer-provisioning) |
| 128 | - Videos: [Admin/Engineer track](/user/videos#adminengineer-track) |
| 129 | |
| 130 | ## Developer / AI agent |
| 131 | |
| 132 | You are extending or modifying CoPilot. |
| 133 | |
| 134 | You care about: |
| 135 | - Architecture + data flows |
| 136 | - Schema changes (Alembic as source of truth) |
| 137 | - Adding a connector safely |
| 138 | |
| 139 | **Start here:** |
| 140 | - [Developer start here](/developer/start-here) |