Clinton Kosh / systems
INDEPENDENT n8n WORKFLOW DEMONSTRATION / TECHNICAL ACCOUNT MANAGEMENT

When the workflow fails, customer trust shouldn't.

An automation reliability command center: detect unhealthy jobs, classify business risk, build a defensible response plan, and communicate clearly with the customer. All orchestrated as an importable n8n workflow.

Run interactive incident replay ↗Download n8n workflow JSON

Built by Clinton Kosh for technical hiring evaluation. The browser replay uses fictional incidents. Live production customers, credentials, and provider connections are not exposed.

DEMO / PORTFOLIO HEALTH SYNTHETIC DATA
34/ 100 health score
EnvironmentSELF-HOSTED
Incident priorityP1 / SEV1
Customer update15 MIN
Actual production writes0
01 / LIVE BROWSER REPLAY

Break the workflow. Watch the response.

Select a fictional account. The browser replays the same documented scoring policy used in the downloadable n8n workflow. It prepares a customer update and recommended recovery actions, without contacting anyone or changing infrastructure.

SEV1SYN-001POLICY v1.0

Atlas Health

12.6%
5,100ms
410
Why this was flagged

CUSTOMER COMMUNICATION DRAFT

Prepared only. Human approval required before sending a real customer update.

RECOVERY PLAYBOOK

    66 / 100

    Weighted by failure rate, p95 latency, queue depth, last successful run, and repeat incidents.

    TAM + Incident Commander

    Incident bridge

    SAFE REPLAY

    No production changes. No customer data. No credentials. Recommendations are subject to owner approval.

    02 / ACTUAL IMPORTABLE n8n WORKFLOW STRUCTURE

    Signal to decision to accountable action.

    Nine connected workflow nodes. A manual replay creates synthetic telemetry; subsequent nodes validate, score, route, prepare communication, propose recovery, and emit a machine-readable proof receipt.

    1
    Manual replayTrigger
    2
    Generate telemetrySample fixtures
    3
    Validate + sanitizeInput gate
    4
    Health + SLORisk engine
    5
    Escalation policyOwner + SLA
    6
    Customer statusDraft only
    7
    Safe recoveryHuman gate
    8
    Audit receiptAssertions
    9
    Executive rollupProof payload
    Schematic of downloadable workflow source. Import into n8n to see the actual node canvas; this diagram is not presented as an n8n UI screenshot.
    03 / WHAT A TECHNICAL ACCOUNT MANAGER OWNS

    Engineering context, translated into a customer outcome.

    01 / PROACTIVE HEALTH

    Detect risk before escalation.

    Track SLO deviation, latency, failing automations and queue pressure. Correlate recurring defects into preventable account risks.

    02 / INCIDENT OWNERSHIP

    Keep the response coordinated.

    Identify the technical owner, escalation priority, customer update cadence and safe troubleshooting sequence.

    03 / OPERATIONAL PROOF

    Make each decision reviewable.

    Keep explicit severity rules, non-destructive playbooks, approval checks and an audit receipt that can be revisited after the incident.

    04 / SOURCE AND PROOF BOUNDARY

    Inspect the actual artifact.

    Download the workflow

    An editable n8n JSON with connected Manual Trigger and Code nodes. Run it locally to examine scored incidents and the executive rollup. No credentials required.

    n8n workflow JSON ↗

    What this does not claim

    The public UI is a client-side policy replay, not a live view into a private n8n instance. Fixture counts and scores are illustrative. It does not send messages or perform autonomous production remediation.

    synthetic: true productionWrites: 0 approvalRequired: true

    Technical notes: deterministic risk policy; minimal dependencies; no public unauthenticated webhooks; no private hostnames; real n8n runtime validation should be demonstrated by an execution receipt, not inferred from source alone.