Frequently Asked Questions

Getting Started

What is FlautoPsy?

FlautoPsy is a monitoring tool for AI workflows running on Make.com, n8n, Zapier, or custom APIs. It detects when your LLM prompts, outputs, latency, or cost deviate from normal behavior and alerts you via Slack or in-app notifications.

How does FlautoPsy work?

  1. You send traces (prompts, outputs, latency, cost) from your AI workflow to FlautoPsy
  2. After 10 traces, FlautoPsy learns your baseline — what "normal" looks like
  3. On new traces, we detect drift — significant deviations from that baseline
  4. You get alerts immediately

Learn more in Core Concepts.

What do I need to get started?

  • A Make.com, n8n, or Zapier account (or any HTTP-capable application)
  • An active AI workflow (using GPT-4, Claude, etc.)
  • A FlautoPsy account (free tier available)
  • (Optional) A Slack workspace to receive alerts

How long does setup take?

5-10 minutes. See Getting Started for step-by-step instructions.

Is there a free trial?

Yes! FlautoPsy is free while in beta. See pricing for plans.

Integration & Technical

Which platforms does FlautoPsy support?

  • Make.com — via HTTP module
  • n8n — via HTTP Request node
  • Zapier — via Webhooks by Zapier
  • Custom applications — any HTTP-capable app via our API

See Integration Guides for setup instructions.

How do I send traces to FlautoPsy?

Send a POST request to your webhook URL with a JSON body containing:

  • text
    workflow_id
    — unique identifier for your workflow
  • text
    prompt
    — the prompt sent to the LLM
  • text
    output
    — the LLM's response
  • text
    latency_ms
    — response time in milliseconds
  • text
    cost_tokens
    — total tokens used
  • text
    model
    — model name (e.g., "gpt-4")

See Custom Integration for details and code examples.

What if I have a question about integration?

Check the relevant integration guide:

Or email support@stralocroft.com.

Can I test my integration before going live?

Yes! Run your workflow manually 10+ times and check FlautoPsy dashboard → Traces to see if they're recorded. If traces appear, your integration is working.

Do you support webhook retries?

Currently, we accept traces once. If a trace fails to send, please retry from your workflow platform's integration settings.

Monitoring & Alerts

When do I get alerts?

After your baseline is mature (≥10 traces), alerts trigger when:

  • Your prompt loses key constraints (PROMPT_DECAY)
  • Your output length deviates by > 3 standard deviations (OUTPUT_DRIFT)
  • Your cost increases > 5x baseline (COST_SPIKE)
  • Your latency increases > 5x baseline (LATENCY_SPIKE)
  • Your workflow structure changes (TOPOLOGY_CHANGE)

See Drift Detection Types for details.

How do I receive alerts?

In-app: Alerts appear in your FlautoPsy dashboard under Alerts → Active

Slack: (Optional) Connect your Slack workspace in Settings → Slack Integration and choose which alerts to send.

Email: Coming soon.

Can I customize alert thresholds?

Yes. Go to Settings → Alert Tuning to adjust sensitivity for each detector or disable specific detectors entirely.

What if I want fewer alerts?

Increase the alert thresholds in Settings → Alert Tuning. For example:

  • Increase COST_SPIKE threshold from 5x to 10x
  • Increase LATENCY_SPIKE threshold from 5x to 7x
  • Disable PROMPT_DECAY if you intentionally change prompts frequently

What happens if I get an alert?

  1. Check your FlautoPsy dashboard for details (which metric drifted, by how much)
  2. Investigate your workflow (did something change? Is the LLM provider having issues?)
  3. If it's a false alarm, adjust thresholds in Settings
  4. If it's a real issue, fix your workflow and restart drift detection

Data & Privacy

How is my data stored?

Workflow traces are stored in a PostgreSQL database hosted on Supabase (AWS us-east-1). All data is encrypted in transit (TLS 1.3) and at rest (AES-256).

How long is my data retained?

  • Free tier: 90 days
  • Pro tier: 1 year
  • Agency tier: Unlimited

You can delete traces manually from the dashboard anytime.

What data do you collect?

  • Account info (email, workspace name)
  • Traces (prompts, outputs, latency, cost) — capped at 10,000 chars per field
  • Usage analytics (number of traces, workflows, alerts)
  • Device/browser info (for product improvements)

We do not collect:

  • Payment info (processed by Stripe)
  • IP addresses (except for rate limiting)
  • User behavior outside of FlautoPsy

See Privacy Policy for details.

Is my data encrypted?

In transit: Yes, all data is encrypted with TLS 1.3

At rest: Yes, database is encrypted with AES-256 via Supabase

In API keys: API keys are hashed with SHA-256 in the database and never stored in plaintext

Can I export my data?

Yes. Go to Settings → Export to download all your traces, workflows, and alerts as JSON.

How do I delete my account and data?

Go to Settings → Workspace → Delete Workspace. This removes all traces, workflows, and alerts. The action is permanent and cannot be undone.

Account & Billing

How much does FlautoPsy cost?

FlautoPsy is free during the beta period. See pricing for planned tiers.

When will paid plans start?

TBD. You'll receive 30 days notice before the free tier ends.

What's included in each plan?

  • Free: Up to 100 workflows, 90-day data retention
  • Pro: Unlimited workflows, 1-year retention, custom thresholds
  • Agency: All Pro features + 2-year retention, SLA

Details on pricing page.

Can I change plans anytime?

Yes. Upgrade or downgrade in Settings → Billing. Changes take effect immediately.

What's your refund policy?

We offer pro-rated refunds for monthly subscriptions. Contact support@stralocroft.com within 7 days of purchase.

Do you offer annual discounts?

Yes! Annual plans are 20% cheaper than monthly. See Settings → Billing.

What happens if I exceed my plan limits?

  • Trace limit exceeded: We'll send you an email and you can upgrade
  • Data retention exceeded: Older traces are automatically deleted per your plan
  • Workflow limit exceeded: You can't create new workflows until you delete old ones or upgrade

How do I invite team members?

Go to Settings → Team → Invite Member. Choose a role:

  • Viewer — read-only access to dashboard
  • Editor — can create/modify workflows, adjust thresholds
  • Admin — full access, can manage billing and team

Security

How do you handle security?

  • All data is encrypted in transit (TLS 1.3) and at rest (AES-256)
  • API keys are hashed and never stored in plaintext
  • Infrastructure hosted on Vercel (SOC 2 Type II) and Supabase (SOC 2 Type II)
  • Regular security audits and penetration testing
  • Row-level security on all customer data

See Security page for details.

How do I report a security issue?

Email security@stralocroft.com with details. Please allow 48-72 hours for a response.

Do you store my prompts and outputs?

Yes, we store truncated prompts/outputs (max 10,000 chars) to detect drift. We don't use them to train models or improve our detection algorithms.

You can:

  • Delete traces manually from the dashboard
  • Set data retention to 90 days to auto-delete old traces
  • Contact support@stralocroft.com to request full deletion

Troubleshooting

I sent a trace but it's not showing up

  1. Check the response from your HTTP request — should be HTTP 200
  2. Verify the JSON is valid (use a JSON validator)
  3. Check that all required fields are present (workflow_id, output, latency_ms, cost_tokens, model)
  4. Wait 5-10 seconds and refresh the dashboard
  5. Go to Traces tab to see if it appears

My workflow ID is wrong — how do I fix it?

Go to Workflows → Click your workflow → Settings → Rename Workflow. Or delete the workflow and send new traces with the correct ID.

How do I prevent alerts on expected changes?

If you intentionally change your workflow:

  1. Temporarily disable specific detectors in Settings → Alert Tuning
  2. Run your new workflow 10+ times
  3. Re-enable detectors once the baseline adapts

Or reset the baseline: Settings → Reset Baseline.

Can I test my integration without running real workflows?

Yes! Send test traces using cURL, Postman, or code. See Custom Integration for examples.

What should I do if I disagree with an alert?

  1. Check Drift Detection Types to understand why it triggered
  2. If it's a false alarm, increase thresholds in Settings → Alert Tuning
  3. Contact support@stralocroft.com if you think there's a bug

How is SLA uptime defined?

We aim for 99.9% uptime. This means ≤43 minutes of downtime per month. Maintenance windows are not included. See Status page for current uptime.

What's your uptime SLA?

We target 99.9% uptime during business hours. See Status page for real-time status and historical metrics.

Still Have Questions?

Can't find an answer? Email support@stralocroft.com and we'll respond within 24 hours (beta hours may vary).

You can also: