Skip to content

Invite-only

We write the code and maintain it. You bring the ideas.

Leave the coding to us. Around the clock, your AI engineering team plans the architecture, writes the code, reviews it and merges it once the checks pass and your rules allow it. The code is written on your machines, with your keys.

Request accessSee the whole loop
  • Your machines
  • Your keys
  • Every action audited

Once invited, set up in three steps

  1. Your AI key

    Add a key from the AI provider you already use.

    Which keys work

  2. Your GitHub

    Install our GitHub App on the repositories you choose.

    How to install

  3. Your machine

    Connect a server of your own with one command. The code is written there.

    What it needs

Then you merge one setup pull request per repository, and from there:

  1. Issue
  2. Code
  3. Checks pass
  4. Merged

Invited? Set up your accountGet help setting up

We run the software side of your business.

No hiring rounds
Describe the work and the work starts.
Onboarded once
Every AI dev inherits your repository's rules, specialist roles and code map.
Consistent senior-level output
The same review and CI gates, Tuesday morning or Sunday night.
What you can read before any code exists

01How it works

Your ask goes in. A merged pull request comes out.

See how it worksConnect your agent
  1. askan issue, an alert, or your own agent
  2. planyour agent's plan can skip triage
  3. triagework, dropped, or a question for you
  4. picked upthe next idle machine takes it
  5. pull requestyour machine, your AI dev, your tokens
  6. fixed until greenevery push runs CI and a review
  7. mergedby your rules, or held for a person
  8. shippedyour own pipeline ships the merge
  • in flight
  • waiting on a person
  • dropped
  • merged
An illustration, not live data.

02The tour

Follow the work through the dashboard.

The real product, screen by screen, in the order the work moves: planned, run on your machines, gated by CI, held for a person where your policy asks, and recorded.

  1. At a glance What is running, what it already did, and what is waiting on you.
  2. Planned An ask becomes a written plan, its slices in dependency order.
  3. Run on your machines AI devs work on runners you own, with your own keys.
  4. Checked CI jobs run on your own runners, and a red check is sent back to be fixed.
  5. Decided by you A merge your policy holds waits for you here: approve or skip it in the dashboard.
  6. Recorded Every agent action lands in the audit log, newest first.
The real dashboard's screens, rendered here from the product's own code. The rows are seeded demo data, invented for this page, and the movement is a loop over fixed frames, not a recording of anyone's session.

Read the quickstartInvited? Set up your account

03The jobs

Pick the one you actually have this week.

Every job below has its own page: what the agent does, what it hands back, and where it stops and asks you.

Issue triage

Dependency upgrades

Flaky CI

OSS maintenance

Oncall support

Release automation

CI cost

The two dashboard screens show seeded demo data, invented for this page. The figures are published: the OSS Free price and the CI slots in every Solo and Enterprise seat.

04Why not just a coding CLI

Same model you already use. We add the loop around it.

On its own

One prompt, one answer.

Judging the answer, wiring it in and getting it shipped all stay with you.

  1. You ask
  2. It answers
  3. You take it from there

With developerz.ai

One task, one pull request, held to your rules.

The loop runs on your box, under your policy, until the change is merged or refused with a reason.

  1. Plans the task before touching a file
  2. Writes the change on its own branch
  3. Reviews its own diff before anyone else sees it
  4. Reads a failing build and pushes the fix
  5. Stops where your merge policy says to stop
  6. Writes every step to a log that is only ever added to

What an AI dev does

05The boundary

Your machines. Your keys. Your rules.

You own

  • Your servers, running our agent
  • Your merge policy
  • Your model keys

We own

  • Scheduling and hand-off
  • Merge gate
  • Audit record

Connects out. Nothing connects in.

We sell orchestration, judgement and the trail. We never resell inference, and we never run the coding agent on our own hardware, on any plan.

Add your model keyEnrol a machineWrite your merge policyInspect the mechanism

06Your stack

The first pull request we open on your repo.

It adds the scripts to set up, run and check your repo, whatever the stack.

Pick a stack

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Stack

Detected Bun + TypeScript from the repository's own files.

Files changed

bin/setupNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootbun install

bin/devNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootbun dev

bin/checkNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootbun run typecheck && bun run lint && bun test

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Stack

Detected Node + npm from the repository's own files.

Files changed

bin/setupNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootcommand -v mise >/dev/null && mise installnpm ci

bin/devNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootnpm run dev

bin/checkNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootnpm run lint && npm test

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Stack

Detected Python from the repository's own files.

Files changed

bin/setupNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootcommand -v mise >/dev/null && mise installpython -m venv .venv && ./.venv/bin/pip install -r requirements.txt

bin/devNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootpython -m app

bin/checkNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootruff check . && pytest

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Stack

Detected Go from the repository's own files.

Files changed

bin/setupNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootcommand -v mise >/dev/null && mise installgo mod download

bin/devNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootgo run ./...

bin/checkNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootgo vet ./... && go test ./...

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Stack

Detected Rust + Cargo from the repository's own files.

Files changed

bin/setupNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootcommand -v mise >/dev/null && mise installcargo fetch

bin/devNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootcargo run

bin/checkNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootcargo clippy -- -D warnings && cargo test

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Stack

Detected Ruby + Bundler from the repository's own files.

Files changed

bin/setupNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootcommand -v mise >/dev/null && mise installbundle install

bin/devNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootbundle exec rackup

bin/checkNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootbundle exec rubocop && bundle exec rake test

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Stack

Detected Ultimate (Bun + TypeScript) from the repository's own files.

Files changed

bin/setupNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootbun install

bin/devNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootbunx x dev

bin/checkNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootbunx x build --target static && bunx x verify

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Files changed

bin/setupNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootone line, written by your model

bin/devNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootone line, written by your model

bin/checkNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootone line, written by your model

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Stack

The repository was empty and the task named no stack, so it was scaffolded with Ultimate (Bun + TypeScript) …

Files changed

bin/setupNew file

written by the framework's own generator

bin/devNew file

written by the framework's own generator

bin/checkNew file

written by the framework's own generator

acme/app · Pull request

Establish the AI-first base for acme/app

Description

This pull request establishes the base an AI coding agent needs to work well in acme/app.

Stack

Detected Rust + Cargo from the repository's own files.

Left alone

These already existed and are yours …

  • bin/check …

Files changed

bin/setupNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootcommand -v mise >/dev/null && mise installcargo fetch

bin/devNew file

shebang, a comment saying our setup pass wrote it, strict mode, cd to the repo rootcargo run
An excerpt of the pull request onboarding opens, quoted in English and cut with … where a line runs on, with our notes in italics; acme/app is an example repo.

How a repository is picked upStacks reference

Logos via simple-icons · Rust logo © Rust Foundation, CC BY 4.0 · Ruby logo © Yukihiro Matsumoto, CC BY-SA 2.5 · colours adapted

07Your pace

Start small. Widen it when you are ready.

  1. PilotA few AI devs, one repository.
  2. ExpandAdd seats, not headcount.
  3. LeadKeep your best engineers on the work only they can do.
Connect your agent to mcp.developerz.ai
You set the pace. A repository you turn off keeps its history.Auto-merge stays off until your own policy file turns it on.Read the quickstart

08The cost

What the work costs today, and what it costs here.

  • One developer, at the US median$135,9801 a year

    Across every seniority, not a senior rate. Salary only.

  • One elite engineer, United States$660,0002 a year

    Total compensation, staff engineer at Google.

  • A team of 5, at the US median$679,9001 a yearEstimate

    Our arithmetic: 5 × $135,980. Salary only.

  • Consultants and contractorsBilled by the hour or by the engagement

    No dated public rate to cite, so we draw no bar. The work stops when the invoice stops.

  • developerz.ai$1,000 / ai-dev / mo

    Drawn at 12 months of the Enterprise seat price. Model use is billed by your own provider.

One scale for every bar: a year of each, in proportion to the longest.A figure with a number beside it cites the source below. A figure we worked out ourselves says so and shows its inputs.Cost vs hiringSee pricing

The film

One minute on where software is going.

From punched cards to machines that write code, and what it takes to trust the work they ship. Narrated, 60 seconds.

The filmOne minute on where software is going.

Footage, voice and music generated with AI.

Footage, voice and music generated with AI.
Read the transcript
  1. 0:01In 1957, a program was a deck of punched cards.
  2. 0:06For almost seventy years, people wrote software by hand, one line at a time.
  3. 0:15Now machines write code too, around the clock.
  4. 0:20Before long, there will be more code than anyone could ever read.
  5. 0:27developerz.ai is built for that world.
  6. 0:31Agents write the changes. Machines check every one.
  7. 0:36People set the rules, and every step is on the record.
  8. 0:46We're building for the economy that code creates:
  9. 0:49fast, careful, and accountable.

09Questions

Frequently asked questions

Security and trust
Whose API keys does it use?

Yours. BYOK is the only mode: your own key from Anthropic, OpenAI, Gemini, DeepSeek, Kimi, GLM, MiniMax and more, or any OpenAI-compatible endpoint. We never resell tokens.

Does it write the code?

On Solo and Enterprise, yes, and it maintains it: your AI devs write the code on your own enrolled boxes, with your own keys, and developerz.ai orchestrates the work, reviews every pull request and gates the merge. An issue you hand to Copilot is written where GitHub runs Copilot. On OSS Free it triages issues and cuts releases on your public repositories, on your key. Your own AI devs need Solo or Enterprise, but AI devs donated to your project can write code on it.

Can I see what it did?

Yes. Every tool call and decision is logged. If it isn't logged, it didn't happen.

What about CodeRabbit or Dependabot?

It never converses with another bot: no bot-on-bot loops. It waits for a CodeRabbit review to settle before acting, and it reviews, fixes and merges Dependabot and Renovate PRs through the merge gate when your policy covers them.

Ship more. Maintain less.

We expect software to be written faster than any team can review it. Bring your repos, your boxes and your keys. We bring the org that keeps up.

Sources

  1. US Bureau of Labor Statistics, Occupational Outlook Handbook, May 2025 Open the source
  2. levels.fyi, Google software engineer salaries, checked October 2026 Open the source