01
Summary
Skippy is deployed entirely inside your AWS account. Sailor does not operate a multi-tenant observability backend and does not retain your traces, investigations, or feedback on our infrastructure.
02
Deployment boundary
- Skippy runs on your infrastructure — never on Sailor-managed compute
- All application state lives in your AWS account
- You control network access and who can reach your systems
03
Observability & telemetry
LLM traces, token usage, cost attribution, and infrastructure metrics are stored in your account — not ours.
- Operational data never leaves your account
- We do not operate a central trace or analytics store for customer workloads
- Sailor does not proxy, copy, or query your telemetry
04
Feedback & audit
- User feedback stays in your account
- No feedback is sent to a Sailor-operated analytics service
- Investigation audit trails remain under your retention policies
05
LLM inference (hybrid mode)
In air-gapped deployments, inference stays fully inside your network. In hybrid mode, Skippy may route inference to a managed LLM to reduce compute cost.
- Inference uses your credentials and your cloud account
- Sailor does not log prompts, responses, or token usage on our side
- Traces from those calls still never leave your account
06
What we never do
- Store customer traces, logs, or investigation data on Sailor infrastructure
- Run a shared observability backend across customers
- Phone home with usage analytics from Skippy deployments
- Retain LLM prompts or responses outside your account
- Require outbound telemetry for Skippy to function
07
Sailor website & sales
This page describes the Skippy product security model. The Sailor marketing site may use standard web analytics (see our Privacy Policy). That is separate from Skippy deployments in your account.
Questions?
Happy to walk through architecture diagrams, data flows, or compliance requirements on a call.
sid@awsistoohard.com