Check the cost breakdown in subcontractor quotes
AI reads labor, materials, and other items in subcontractor quotes and flags missing costs and rates below the public works design labor rate. Staff decide.
- Runs on your own PCs and servers
- Use Azure OpenAI, Amazon Bedrock, Google Vertex AI, or an in-house model
- Its engine is Descarty's open-source Dagu (GPL-3.0)
The problem today
Since Japan's amended Construction Business Act took full effect in December 2025, contractors should try to issue quotes that itemize labor, statutory welfare, and other costs, and those ordering the work should try to take them into account. Yet every subcontractor accounts for health and safety costs and materials in its own way. Staff read each quote and compare its labor rates against the table of public works design labor rates by hand.
The workflow in Kitewell
- Take in quotes (PDF, Excel) that subcontractors send by emailEmail
- AI reads trade, quantity, labor rate, productivity rate, materials, statutory welfare, health and safety costs, and retirement fund contributions, and lists themAI judgment
- Compare with the design labor rate for the site's prefecture and trade, and extract lower rates and missing cost itemsAutomatic
- AI drafts the points to confirm and a question to the subcontractorAI judgment
- Staff review the results and the draft, and decide whether to send itApproval
- Save the initial and final quotes by project and record them in the registerAutomatic
What AI does
Reads the breakdown from quotes in any format and shows where in the quote each value came from. For rates below the design labor rate and missing cost items, it drafts the points to confirm.
Where people decide
Staff judge whether productivity rates and costs fit the site, whether to ask for a revision, and which company gets the contract. AI never decides whether a quote is acceptable.
Prove Kitewell on your own systems
Try it free for 14 days on your own PC. To see it working first, ask for a 30-minute demo. There is also a one-month onboarding (¥500,000 before tax), where we build your first job together.
What to check when you try this work
- Whether the breakdown is read correctly from your main subcontractors' quote formats
- Whether quotes state labor rates on the same basis as the design labor rate (per 8 hours, excluding the employer's share of required costs)
- Whether the rate year, prefecture, and trade used for comparison can follow your purchasing procedure
- Where initial and final quotes are kept, and how they link to each project
How the onboarding runs
- Week 1Requirements and success criteria
- Weeks 2 to 3Build
- Week 4Verify on production screens
- Final dayReport
One-month onboarding
- Length
- One month
- Cost
- ¥500,000 (tax excluded). Your staff build, with the developer beside them
- If you go on to a license
- The onboarding fee is credited to the first year's contract
- Without the onboarding
- At Team scale, start on Team monthly (¥14,800 per computer a month, tax excluded), and cancel in any month
You can try it first on your own PC, in the free 14-day trial.
Frequently asked questions
- Does AI judge whether a labor cost is markedly low?
- No. AI points out rates below the public works design labor rate and missing cost items. The government's reference values assume standard site conditions, and staff judge whether a productivity rate fits the actual site. Staff also decide whether to ask the subcontractor for a revision.
- Can it compare competing quotes?
- Yes. It can line up each company's breakdown under the same items. MLIT's operating policy expects those ordering work, even when choosing the lowest total, to check that labor and other costs are properly estimated and not to choose a quote that falls markedly below them. The comparison is material for that check.
- What should we do with a quote whose labor cost is below the standard?
- MLIT's operating policy expects the ordering party to first ask the contractor what it intended, and, if the quote seems improper, to report it to MLIT's hotline or similar. It also explains that merely receiving such a quote does not place a specific duty to act on the ordering party. Staff decide how to respond.
- Do our documents and screens leave the company?
- Kitewell runs on your own PCs and servers, and Descarty never receives your business data. You choose the AI model: a cloud API, or an in-house or local model with an OpenAI-compatible API. With an in-house model, no data leaves the company. With a cloud model, what the AI reads is sent to the model's provider.
- What if the AI misreads something?
- A person checks what the AI produced before the work goes on. In a sheet, every value the AI reads shows the quote it came from, and rows below your confidence threshold go to a person. Reviewers can send work back with notes on what to fix, and who approved what is kept in the run history.
- What happens when a screen changes or a run stops?
- When a screen changes, the AI reads it again, relearns the step, and carries on. The relearned step is saved, and later runs repeat it. If a run still stops, ask the assistant right there, and it finds the cause and proposes a fix. A proposal is not saved until a person applies it.
- What happens to the RPA and systems we use today?
- Keep them. Kitewell operates screens the way a person does, so it works with systems that have no API. Move the robots that keep stopping and the work you gave up automating because it needs judgment, and leave fast, fixed processing in your current RPA.
- What does it cost?
- ¥14,800 per computer per month, or ¥148,000 per computer when paid yearly, before tax. A computer costs the same whether it builds or runs. You can try it free for 14 days. The one-month onboarding, where we build your first job together, is ¥500,000 before tax, credited to the first year if you go on.
Related use cases
Check safety documents from subcontractors
AI reads subcontractors' worker rosters and certificate copies and finds expired or missing licenses, health checks, and insurance. Staff check requests first.
Read and cross-check import documents before filing
AI reads the invoice, packing list, and B/L and cross-checks goods, quantities, weights, and values. A customs specialist resolves mismatches before filing.
Collect tender notices and pick the ones to bid on
Collect tender notices daily from government portals and municipal sites, and have AI judge which ones you can and should bid on. Sales makes the call.
See this work in a demo
A 30-minute demo shows this workflow running. The 14-day free trial lets you check it on your own systems.