Track placement retention and fee refunds
Confirm with employers at set points up to six months after hire that placed people stayed. For early leavers, draft the refund and log the six-month count.
- 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
Fee-charging placement agencies in Japan must publish online how many of the people they placed in permanent jobs left within six months, not counting dismissals. To confirm that number, the agency must survey the employers. Where the agency has a refund policy, the amount returned depends on how long the person stayed. Staff email each employer to confirm who is still employed and work out refunds against the policy by hand.
The workflow in Kitewell
- Each week, extract placements reaching 1, 3, or 6 months after hire from the core systemWeb
- AI drafts an email to each employer to confirm continued employmentAI judgment
- Staff check the drafts and approve sendingApproval
- Send the emails to employers and receive repliesEmail
- AI reads whether the person stayed or left, the leaving date, and whether it was a dismissal, and calculates a refund draft from the policyAI judgment
- Staff approve the refund and the record for the six-month count, and update the placement recordApproval
What AI does
Drafts an email to each employer to confirm continued employment. From the reply, it reads whether the person stayed or left, the leaving date, and whether it was a dismissal, and proposes a refund under the policy. Replies it cannot read are passed to staff.
Where people decide
Staff decide whether to refund and how much. Where applying the policy is unclear, such as the reason for leaving, staff check with the employer.
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 hire dates, whether the contract is permanent, and refund terms can be retrieved from your core system
- Whether the refund policy (periods and refund rates) is reflected correctly in the calculation
- Whether the timing and wording of employer checks fit your contracts with employers and how you work
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
- May the six-month count be taken from refund cases?
- Japan's Employment Security Act regulations say that an agency with a refund policy need not survey employers if it counts the people who triggered a fee waiver under that policy. Decide which method to use before the PoC. Either way, calculating refunds needs to know whether and when the person left.
- 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 and track dispatch period limits
Site limits come from the client's notice; per-person limits run from the worker's start in the unit. As either nears, AI drafts a check that staff review.
Write job postings for each job board
From a client's job request, AI drafts postings in each job board's format. Staff check them, then publish through each board's official route.
Total client timesheets and watch overtime caps
Collect dispatched workers' hours from each client's attendance system and find people near the overtime cap. AI drafts the note to the client; staff check it.
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.