Service Level Agreement
Clear expectations for response times and support. These are baseline commitments — specific terms are agreed in your engagement contract.
Response Times
How quickly I respond depends on severity. All times are business hours, Jerusalem time (UTC+2).
Production Down
Your AI system is completely unavailable or producing incorrect outputs that directly affect business operations.
Degraded Performance
System is functional but with significant issues — slow responses, partial failures, or noticeably reduced output quality.
Non-Urgent Issue
Something isn't working as expected but there's a workaround, or it's affecting a non-critical workflow.
Enhancement Request
Feature requests, configuration tweaks, prompt refinements, and other non-time-sensitive improvements.
During the Build Phase
Communication patterns while your system is being developed.
Weekly Updates
Written progress updates every week covering what was built, what's next, and any blockers or decisions needed.
Demos at Milestones
Live walkthrough at each major milestone — so you can see the system working and give feedback before I move on.
Async Communication
Email or your preferred channel for day-to-day questions. I aim to respond within 1 business day during active build phases.
During Support Retainer
What to expect once your system is live and under ongoing support.
Uptime Monitoring
Continuous health checks on deployed systems. I'm often aware of issues before you are — and working on them.
Incident Response
When something breaks, I follow the severity framework above. Critical issues get same-day attention.
Monthly Reviews
Written summary of system health, work performed, and recommendations. Delivered by the 5th business day of each month.
Escalation Path
If a response SLA is missed, or if you need to escalate urgency, contact me directly via email or phone. Escalation details are in your support agreement.
Business Hours & Availability
I'm based in Jerusalem, Israel — my working week and calendar reflect that.
Work Week
Sunday through Thursday. Friday and Saturday are offline. Response SLA timers only count business days within this schedule.
Holidays
I observe Jewish and Israeli national holidays. I'll share my holiday calendar at the start of each engagement so you know when I'm unavailable.
Planned Offline Time
Vacations and extended offline periods are communicated at least two weeks in advance. Critical coverage arrangements are discussed in your support agreement.
What's Not Covered
Transparency about the boundaries of these commitments.
- Third-party outages — if OpenAI, your cloud provider, or another external service goes down, that's outside my SLA scope. I will help coordinate response, but can't guarantee their uptime.
- Scope changes — new features or significant architectural changes during an active build are scoped and quoted separately.
- Infrastructure you manage — if you host your own servers and they go down, response time applies to the AI system layer, not your infra.
- After-hours emergencies — standard SLA times are business hours. After-hours emergency support is available by arrangement in your support agreement.
- Systems I didn't build — supporting third-party systems requires an onboarding assessment before SLA terms apply.
Questions about support terms?
Get in touch to discuss how these commitments apply to your specific engagement.