How I work
You deal with me from the first e-mail to maintenance. The numbers first, then the decision, and the code is yours from the start.
Four stages of working together
The bar length is time, and each stage has its price. After the audit you can stop at the report and implement the changes with your own team.
Call
Week 1.
An e-mail or a short call. You show me the bill or the architecture.
free
Architecture and cost review
Weeks 1–2: audits S and M take 5 to 10 working days, audit L 2–3 weeks.
What to change in the architecture, deploys and costs, with the effort and the saving where there is one. Read-only access is enough.
from EUR 690
Implementation
From week 3, length depends on scope.
The changes from the report, or a new environment next to the old one, in Terraform, with a rollback plan. A move to Hetzner or OVH is priced in the exit plan, from EUR 1,850.
EUR 60 per hourfor changes in AWS
Maintenance
Monthly after that.
Updates, backups with restore tests, alerts and a monthly report.
from EUR 190 per monthno annual lock-in
Security and access
- Read-only access to start with
- For the audit, a read-only IAM role that cannot see your data (the policy is on the audit page), or an export from Cost Explorer, is enough. I only get write access for the implementation.
- My own access, no shared logins
- I get an IAM role that I assume from my own AWS account (with an ExternalId), or a named user with MFA. No long-lived access keys and no shared passwords. I make changes through pull requests, and CloudTrail shows each of them. When the work ends, you remove the role with one click.
- Secrets kept out of the code
- Passwords and keys live in Secrets Manager or a password manager, not in the repository and not in e-mails.
- NDA and a data processing agreement
- I will sign an NDA before the first call and, if I process personal data, a data processing agreement too.
Nothing depends on me alone
- The code is in your repository
- The infrastructure is described in Terraform and Ansible from day one, in your repository, not mine.
- The accounts are yours
- Provider accounts, domains and passwords belong to you. You pay the providers directly.
- Runbooks for incidents
- The key procedures are written down, so that any competent engineer could take the infrastructure over without me.
Me, a large consultancy or a full-time DevOps hire
Each option makes sense in a different situation. I compare ways of working together, not specific firms. As of 23 September 2026; the terms of a particular offer may differ.
| CODELEVEL | Large AWS consultancy | Full-time DevOps hire | |
|---|---|---|---|
| Cost | Public prices, net: hourly work EUR 60 per hour, maintenance EUR 190 per month with 1 hour of work included, audit from EUR 690. You pay for the part of a full-time job you need. | Quoted per project after a scoping exercise. | A full-time salary plus the employer’s costs, equipment and paid leave, whether or not there is work that month. What that costs depends on your country and market. |
| Who does the work | Me, the same person from the audit to maintenance. | The firm’s team; the firm decides who is on it. | Your employee, working only for you. |
| Start | Audits S and M: 5–10 working days from receiving read-only access; audit L: 2–3 weeks. | As scheduled in the offer. | Recruitment, then the candidate’s notice period with their current employer, often one to three months. |
| Availability | Business hours in the same time zone as Berlin (CET/CEST), critical incidents in the evening too, until 22:00. No night on-call. | As per the contract; some firms offer round-the-clock on-call. | Working hours; on-call as per the contract and working-time rules. |
| Cover | One person. While I am on holiday or ill, the work waits or your team takes it over, helped by the code and runbooks in your repository. | Several people on the team can stand in for each other. | One person. Leave or a resignation is a gap to plan for. |
| Contract | Maintenance: a monthly contract with one month’s notice. Audit and plan: a contract for a single job. | Per project or a framework agreement, as per the offer. | An employment contract with the notice period your local law sets. |
| Ties to cloud providers | I do not resell cloud providers’ services and take no commissions from them. | The firm may take part in a cloud provider’s programmes. | Not applicable. |
| Best choice when | Your application runs on AWS, another cloud or your own servers and you need specific changes or ongoing care for less than a full-time job. | You need round-the-clock on-call, a large team or several projects at once. | The infrastructure needs full-time, day-to-day work. |
As of 23 September 2026. CODELEVEL prices are net of VAT, in euro, and listed on the pricing page.
When I respond
Monitoring and alerts run all the time. I respond during business hours, in the same time zone as Berlin (CET/CEST), and take critical incidents in the evening too, until 22:00. I do not run a night on-call. Exact response times depend on the maintenance plan, from EUR 190 per month.
Plans and response timesA recommendation to paste
If you liked working with me and someone asks about DevOps or AWS costs, you can paste this paragraph.
Jacob Bińczyk (CODELEVEL) is an independent DevOps engineer, AWS Certified Solutions Architect – Professional. He designs and rebuilds AWS infrastructure in Terraform, runs fixed-price AWS cost audits and, where it makes sense, moves infrastructure to Hetzner or OVH. He works alone, replies within 1 business day, and his prices are public: codelevel.pl/en/pricing/
Start with your bill
Tell me where your infrastructure runs and what you want to change. I will reply with questions, or straight away with a proposal.
Or email jacob@codelevel.pl.
I reply within 1 business day.
From a case study: a SaaS platform I built on AWS has 99.99% availability over a year. See the case study