Growth
Stop firefighting
₪15,000 / month$5,000 a month, before VAT
- One AWS account
- One or two environments
- Up to ~10 services
- Dev team of eight or fewer
- Planned work
- One active workstream, planned together
- Cadence
- A strategy call every two weeks
Pricing
What this costs
Monthly, cancellable with 30 days notice. Prices exclude VAT.
What separates the tiers
How much infrastructure I carry, and how much of my week you get. Every tier covers the same services. Tiers differ by infrastructure scope, planned-work capacity and review cadence. There are no change orders.
You get me, and infrastructure your team can run without me.
Your whole engineering team shares a Slack channel with me. What I build is written down so your team can act when I am not there: every alert links to a runbook, and every change is a pull request that can be reverted. If production is down in my hours, it comes first on every tier — the queue below is for planned work.
Working hours are Sunday–Thursday, 09:00–18:00 Israel time. There is no coverage outside them, and none while I am on leave — leave is announced ahead, with the open work written up.
Scale
Resilience and depth
₪30,000 / month$10,000 a month, before VAT
- Multi-account organisation
- Three environments
- ~40 services or one Kubernetes platform
- Dev team of up to 25
- Planned work
- Two parallel workstreams, scheduled first each week
- Cadence
- Weekly sync and architecture deep-dive
Enterprise
Compliance and scale
from₪48,000 / monthfrom $16,000 a month, before VAT
- Multi-account, multi-region
- Regulated or high-load workloads
- Platforms where an outage is a business event
- Taken on subject to fit
- Planned work
- Your roadmap leads my week
- Cadence
- Weekly sync plus quarterly roadmap
- Also
- Quarterly live failover test
In every tier
- Terraform for all infrastructure, reviewed through pull requests
- CI/CD pipelines for ECS, EKS or serverless
- Monitoring, alerting and on-call runbooks
- Security baseline — IAM least privilege, WAF, encryption at rest and in transit
- Continuous AWS cost optimisation
- Direct Slack access for your whole engineering team
- A direct line to me when production is down, in working hours
Also in every tier, when you need it
- SOC 2, ISO and FedRAMP readiness — Logging, access control and the evidence your auditor will ask for
- Disaster recovery — Design, documentation and live failover testing
- Migration and replatforming — Into AWS, or between compute models
- ML and GPU workload tuning — Cost and throughput on high-load inference
Work this size is why a client is on a higher tier. It is never invoiced twice.
Invoiced in shekels, plus VAT. Dollar figures are a guide at ₪3 = $1, not the invoice.
Straight answers
Questions people ask.
- How fast can you start?
I work with a few companies at a time, so the start date is agreed on the call rather than promised on a page. From the first week, your most urgent problem is the one I am working on, and what I find is written down for you to keep.
- What cloud platforms and tools do you work with?
AWS, GCP and Azure. AWS is where my depth is — I’ve run it since 2018 — and it is where most of my clients live; the other two I run when a client brings them, in Terraform or Pulumi, the same way. Beyond the clouds it is the usual serious stack: Kubernetes, GitHub Actions or GitLab CI, ArgoCD, Grafana, Prometheus, Datadog, and the AWS inference services for teams shipping models. The full list is on the How I work page. If your centre of gravity is somewhere unusual, I’ll tell you on the first call whether I’m the right person.
- Are there long-term contracts?
No. Monthly, cancellable with 30 days notice. I would rather earn the next month than be owed it.
- Can you help with SOC 2, HIPAA or FedRAMP?
Yes, and it is included rather than quoted separately: logging, access control, evidence collection and the infrastructure controls your auditor will ask for. FedRAMP is the same work taken further: the NIST 800-53 controls on the infrastructure side, FIPS-validated endpoints, and the continuous-monitoring evidence your assessor reviews every month.
- What if we want to bring it all in-house?
Planned for from day one. Everything lives in your repositories with READMEs and architecture diagrams, and I run a structured handover to your team.
- You're one person. What happens if you're unavailable?
Outside my working hours, and while I am on leave, nothing — I do not sell coverage I cannot honestly provide, and there is no response-time guarantee. What you have instead is built for exactly that: infrastructure in code, every alert linked to a runbook your team can follow, and every change as a pull request that can be reverted. Leave is announced ahead, with the open work written up. In working hours, production down gets a direct line to me on every tier.
Not sure which tier you are?
That's a two-minute conversation, not a sales process. Tell me how many accounts and services you run and I'll tell you.
A few companies at a time