VIENJA
Get a quoteFree 30-min call
Back to case studies
AI ProjectsAn AI product built by iPlus

Hearth — Running AI-DLC with Claude Code

A workspace that helps a software team turn business intent into working product through AI-DLC: the AI plans and writes code, and people approve at every point that matters.

Book a 30-minute callSee inside the product

On sample data · measured again before and after go-live

from request to production, one unit
9 days−79%
Before: 6 weeks
test coverage
86%+25 pts
Before: 61%
change failure rate
6%−16 pts
Before: 22%
How it works

How Hearth works

The problem

  1. The team gets a request: dealers order online, with volume discounts, and orders above the credit limit are blocked. The details are scattered across chats and meetings.
  2. Each engineer uses AI differently; code comes fast but quality varies from one pull request to the next, and reviewers are overloaded.
  3. From request to production takes about 6 weeks, and defects still slip through.

What it does

  1. 01The AI reads the request and asks about what is unclear: “Above the credit limit, block the order or hold it for approval?” The team answers once, and the request becomes stories with acceptance criteria.
  2. 02Several Claude Code agents write code and tests in parallel, in short rounds; a pull request merges only after an engineer approves.
  3. 03Every feature traces from request to story, pull request, test and release.
  4. 04Team decisions and lessons from incidents are stored, so the AI does not repeat an old mistake.
  5. 05The method follows AI-DLC as published by AWS (iPlus is not affiliated with AWS).
Real product, sample data

Inside Hearth

01/10

The AI-DLC lifecycle on one screen: three units at three stages, each gate showing who approved and when

Rollout

Putting it to work

Rollout with iPlus

  • Weeks 1–2 · AssessmentChoose a pilot repository, connect the code host and ticket tool, and write CLAUDE.md with the team.
  • Weeks 3–8 · BuildA 2-day workshop on your own codebase, the pull request gate and the quality and security gates, then a live run with a team of 6–8.
  • Monthly · AI AdvisoryAfter 4 weeks of support, optional, to update rules and skills and re-measure before and after.
  • You provideAccess to the pilot repository, one technical contact and a PO for the requirement sessions.

Running cost

  • SetupA fixed price covering a 2-day Claude Code workshop, an AI usage policy and Hearth setup.
  • Claude seatsAbout $100–200 a month per engineer (Team or Max plan), paid to Anthropic directly.
  • HostingYour own cloud or a small server; you can start with a few engineers.

Where the AI gets its information

  • Project onlyThe code, tickets, CLAUDE.md, design notes and the decision log.
  • Kept outAPI keys, personal data and copies of the production database.
  • LoggedWho approved, when, and which change passed which gate.
  • No trainingYour code and data are never used to train models.
Built withClaude CodeClaude Agent SDKMCPGitLab APIGitHub APIJira APITerraformNode.jsPostgreSQL

Frequently asked questions

Short answers to what companies usually ask before they start.

Ask anything else in 30 minutes
What if the AI writes something wrong?
Every change passes tests and a human reviewer before merge. If it is wrong the pull request is sent back and the lesson becomes a rule.
Does the team need training?
Yes: a 2-day workshop on your real codebase, then 4 weeks of support. After that, engineers read diffs and approve pull requests, as they do today.
Does it work with the tools we already use?
Hearth connects to GitLab or GitHub, Jira or Backlog, and Slack or Teams. Other tools are checked during the AI Workflow Assessment.
Is our source code safe?
Your code stays in your own repository; iPlus keeps no copy. It is never used to train models.
What does it cost each month?
A Claude seat per engineer (about $100–200 a month, paid to Anthropic directly), small hosting, and AI Advisory if you choose it.
How long until it is running?
Usually 4–6 weeks for a live setup with a team of 6–8 people. The exact schedule is fixed in the contract.
What happens if we stop?
The code, CLAUDE.md, requirement documents and decision log all sit in your repository, so the team keeps them after stopping.
Can we start small?
Yes. Start with one repository and one team, and widen it when the before-and-after figures show it works.

Related projects

A faxed order is read and matched 3 of 3; the price is 12% over the old quote, so Orderly holds it for review.
AI Projects

Orderly — AI Agent for Purchase Order Entry

Orderly reads purchase orders from fax, email and PDF, finds the right part numbers and creates the order in kintone. Unusual orders wait for a person.

The office floor: seven AI agents work beside staff, and every decision that matters goes to the approval desk
AI Projects

Atrium — An AI Agent Office That Works Beside Your Staff

Atrium is an AI agent office: seven agents sort email and WhatsApp, draft quotes, enter orders and chase receivables, while people approve every step that matters.

Luma answers a braces price from the 2026 price list and books Saturday in the chat, with no front-desk work.
AI Projects

Luma Care — Patient Care Chatbot for Dental Clinics

A chatbot that answers patients 24/7 on WhatsApp, LINE and the website, books on the real dentist calendar and hands clinical cases to a dentist. Every answer shows its source.