Build software 80% faster without losing rigour
We integrate AI tools that read your codebase, architecture and business logic — so generation, review and onboarding all speed up together.
Generic AI tools don't know your codebase.
Engineering time disappears into boilerplate, obscure bugs, PR review and legacy maintenance. Generic assistants help at the margins, but they have never seen your repository.
So they produce code that breaks your patterns and misses your architectural conventions, creating more cleanup than they save. Without your context, AI is just autocomplete.
Generic AI tools have no model of your codebase.
Generated code breaks your conventions and patterns.
Code review takes hours and still misses real issues.
New developers take weeks to become productive.
Technical debt grows faster than you can pay it down.
Context switching between tools breaks engineering flow.
Agents build and test. People release.
Hand over a ticket. Agents take it from spec to a running test environment, with no human in the loop. Nothing reaches production until a person checks it and approves the release.
Agents only · no human in the loop
Human release
Zone one, agents only, with no human in the loop. Step 1, intake: a ticket, an epic or one line of instructions. Step 2, analyse: scope, risk and size. Step 3, spec: an OpenSpec change plus hidden scenarios. Step 4, build: several developer agents work in parallel. Steps 5 and 6 run in parallel on the same commit: verify runs the app against the hidden scenarios, and review checks the change with a fresh context and no developer notes. If either fails, a fix round sends the findings back to build, up to two times. Step 7, merge to main, one change at a time once CI is green. Step 8, the test environment is deployed from main automatically.
Then a release gate. Zone two, human release. Step 9, a person reviews the test environment and approves. Step 10, production: your team releases it.
- Steps 01–06
Agents build and check
An analyst scopes the ticket and a spec writer turns it into an OpenSpec change. Developer agents build it in parallel, then a verifier and a reviewer check the same commit. Failures go back for up to two fix rounds before a person is asked.
- Steps 07–08
Merged and deployed to test
When CI is green, the change merges to main, one at a time. Your pipeline deploys main to the test environment. No one has had to read the spec or the code to get this far.
- Steps 09–10
People release
A person checks the change in the test environment and approves it. Your team releases it to production with your own process. The factory never deploys there.
7090 Open Factory
7090 Open Factory is our open-source software factory for AI coding agents. Give it a ticket, an epic or one line of instructions. A team of agents analyses the work, writes the spec, builds it, tests it against hidden scenarios, reviews it and merges it to main.
Agents checking agents only works when the checks are independent. So the agents that judge the code never wrote it, never see the developers' notes, and no agent may weaken a check to get a pass.
Status: proof of concept. By default it merges without human review, so start on a non-critical repo. You can switch on spec approval or human merge, and hold auth, payments, destructive migrations and infrastructure for a person.
Spec before code
A spec writer drafts an OpenSpec change: proposal, requirements with WHEN/THEN scenarios, design and tasks. It has to pass strict validation before anything is built.
Parallel developer agents
The tasks are split into work packages. Each package gets its own developer agent in its own git worktree, tests first. Each issue ends as one pull request.
Hidden scenarios
The spec writer also writes 3 to 8 holdout scenarios that the developers never see. Only the verifier runs them, against the running app.
Independent review
The verifier and the reviewer start fresh, without the developers' notes. The reviewer checks spec coverage, bugs, security and gate tampering: skipped tests, loosened lint, edited CI.
Merges, never deploys
Changes merge to main one at a time, after CI passes. The factory never deploys, tags or releases to production. That stays with your team.
Works with your coding agent
Built on Claude Code. Codex, Gemini CLI, GitHub Copilot and Cursor read the same Agent Skills format, so they are supported, but not yet tested.
How we solve it
We don't just install AI tools. We build an acceleration layer around your team, your repository and your delivery pipeline.
Codebase-aware AI assistants
Assistants that index your whole repository — architecture, patterns, business logic, standards — so every suggestion arrives with your context attached.
Intelligent code review
Automated review that catches bugs, security issues and pattern violations in the pull request, before they reach production.
Automated code generation
Boilerplate, tests, migrations and feature scaffolds written to your existing conventions, so the first draft is already reviewable.
Team workflow integration
AI wired into VS Code, GitHub, Jira and Slack. Your team gets the speed-up without changing how they work.
Pipeline optimisation
We profile your delivery pipeline, find the steps that cost the most time, and automate them: CI, testing and deploys to your test environment.
Faster developer onboarding
New developers navigate the codebase with AI and get grounded answers about your architecture instead of queuing for a senior engineer.
Every engagement includes
- A software factory on your repos, with a human release gate.
- Codebase-aware AI assistant setup and configuration.
- AI workflow integration with the tools you already use.
- Automated code generation pipelines.
- Automated code review inside your pull request flow.
- Delivery pipeline audit and optimisation.
- Team training and ongoing enablement.
- CI/CD acceleration.
- Documentation generated from your code.
Ready to build faster?
A free 30-minute call. We'll take one slow workflow and tell you what an AI-accelerated pipeline would change.
Related solutions
AI App Modernisation
Your legacy website or application, rewritten in this software factory, with before-and-after numbers to prove the improvement.
Learn moreSpec-Driven Development
Turn product ideas into specs an AI can build from, so generated code lands closer to right the first time.
Learn moreCustom Software Development
Full builds delivered on this same accelerated workflow — web apps, mobile apps, APIs and enterprise systems.
Learn moreAI Training & Enablement
Workshops that get your engineers using these tools well, not just installing them.
Learn more