Skip to content

For dev teams

AI-native SDLC: AI in your dev team with rules and checks

An AI-native SDLC is a development lifecycle where AI agents write code to shared rules and every change passes automated checks before a person sees it. It is how we work ourselves: AI writes all our code, and we have shipped 15+ production projects since 2023.

What it looks like in practice

A development team wanted to turn AI from theory into daily practice. In a two-day workshop and hackathon it set up a shared flow: brief, generate, test, fix. In two days it built prototypes and left with a playbook that became part of its daily work.

Workshop from brief to production

What changes for you

Review stops being the bottleneck

Harness engineering: shared rules for agents (AGENTS.md, CLAUDE.md) and gates in CI: tests, lint and AI review. The agent gets feedback from machines before a person opens the PR. Our public starter shows what that looks like.

Claude Code production starter on GitHub

The agent works in a loop with checks

Loop engineering: the agent goes from task to plan, code and tests up to a PR, and fixes its own work on every failed check. This is how we work too: today AI writes all our code, and Filip is accountable for production.

How we work ourselves

The agent knows what the team knows

Context engineering: conventions, architecture and decisions live in the repository, not in heads and on laptops. A talk for a software team planned for 90 minutes ran to 124, because everyone had their own rules and nothing shared.

Agent = model + harness

What people come to us with

A CTO whose team already uses AI“The team has Copilot and Claude, but pull requests bring more noise than results.”

How we solve it: We set up the rules, review and automated checks so AI code reaches production safely.

Example: A workshop for a software team: real backlog work with AI agents, from the brief to a working prototype. Case study

A CTO with a growing backlog“We are not shipping fast enough and the backlog grows faster than we close it.”

How we solve it: A senior joins your repository and pull request flow from week one.

Example: Filip delivered AI features for a SaaS that 60,000+ people used daily. Case study

How we work

The hard part is not switching AI on, it is setting it up so the team gets faster. We build three layers right in your repository: harness engineering (rules and gates), loop engineering (the agent loop with checks) and context engineering (what the agent knows about your code). On your real backlog, not a demo project, and all of it stays with your team.

We get back to you within two business days, and together we pick what to fix first.
Instead of a quote you get a list: what will be done and how you can check it.
The first changes go through your PR flow in week one.
Then small steps every week or two.
Senior judgment, AI as the method.
This is how we have shipped 15+ production projects since 2023.

When we don't recommend it

  • You are a team of 1-2 developers. A short AGENTS.md you write in an afternoon is enough.
  • You have no tests or CI yet. Start there; without them the agent has nothing to check against.
  • Your team cannot keep up with the backlog itself. That is what Senior developer for your team is for.
  • You only want slides or a talk on AI for leadership. We do not make them.

Results

Workshop, Developer Training

Prompt to Production: AI for Software Teams

Hands-on workshop for Kontentino's dev team. They learned to ship faster with AI tools. Working prototypes proved the results.

Kontentino

Proof, not promises

The goal is a working business. Software is just the tool.

“Digitaliko built our AI image generator, and our users love creating social media visuals effortlessly.”

Hana Nováková
Hana Nováková
CEO, Kontentino

What month one looks like

  • Week one: repository access, one real task from your backlog and a written list of what will be done after month one. The guarantee is measured against it.
  • Every change goes through your PR flow and your review. Contact at least once a week, directly with Filip.
  • End of the month: merged PRs and what stays in your repository: rules (AGENTS.md, CLAUDE.md), hooks and CI checks.
  • Monthly subscription: one stream of work in your backlog, weekly iterations, directly with Filip. Need two projects at once, for example the backlog and setting up an AI-native SDLC? We agree on it.

What it costs

Try the first month. If we miss what we wrote down, we refund 50 %. Monthly subscription from €3,980 excl. VAT, pause or cancel any time. Or a fixed price per project: tell us what you need and we send you a quote and a list of what we deliver. Code, data and infrastructure owned by the client from day one.

Month one starts with a written list: what we deliver and how you can check it. The 50 % money back is measured against it.

Only your own hosting, AI model usage and the AI tool licences for your team. You pay them directly to the providers, not through us.

After launch you get 30 days of support. Then you pause the subscription. When something needs changing or adding, you resume it and we carry on. After you cancel, the software, data and infrastructure stay yours. When you leave, we fix the bugs from the last month.

from€3,980/ month excl. VAT

Let's ship what's holding your product backOr a 15-minute meeting with Filip

Further reading

Engineering
Harness and checks

Vibe coding vs agentic engineering: what's the difference?

Vibe coding accepts AI output that looks finished. Agentic engineering puts the same AI inside a closed loop: a spec, tests, hooks, a review gate and a person who merges. What breaks without the system, and the code that closes the loop.

Harness and checks

Harness engineering in practice: from issue to PR

Harness engineering in practice: the code around an AI agent that decides what it takes, when it is done and who checks it. Our loop from a GitHub issue to a reviewed PR, the code that carries it, and what broke.

Harness and checks

Red-green TDD with AI agents: the test must fail first

AI agents write tests that pass, not tests that prove anything. Our gate runs every new test against the base branch first: if it passes there too, the PR is red. The script, how to add it to CI, and the bug that gave a false pass.

Frequently asked questions

15-minute meeting with Filip
What is an AI-native SDLC?
A development lifecycle where AI agents write the code and the team steers them through rules, checks and review. The difference from "we have Copilot": rules, gates and context live in the repository and apply to everyone on the team the same way.
How much does setting up an AI-native SDLC cost?
The monthly subscription starts at €3,980 excl. VAT: one stream of work, weekly iterations, directly with Filip. Two projects at once we agree on. For a bounded scope we send you a fixed price.
How is it different from Senior developer for your team?
Senior developer for your team is for a team that cannot keep up with its backlog. AI-native SDLC is for a team that has AI but, without rules and checks, is slowed down by it. Often the two combine: the senior ships and sets up the harness along the way.
Does it work with Claude Code, Codex or Cursor?
Yes. We write the rules into AGENTS.md and CLAUDE.md, which these tools read, and the checks run in CI regardless of the tool. The team can switch tools later; the harness stays.
Is it a workshop?
Not only. A workshop can be the start, but the result is rules, checks and an agent loop in your repository, set up on your backlog. When we leave, they stay with you.

Where is AI slowing your team down the most?

Tell us about it. We get back to you within two business days, and together we pick what to fix first.

Let's ship what's holding your product backOr a 15-minute meeting with Filip