Match food product pages to spec sheets
AI reads suppliers' food spec sheets and compares them with the ingredients, allergens, and origin on your product pages. Staff confirm before pages change.
- 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
Online food product pages show ingredients, allergens, and origin taken from the spec sheets suppliers send. Spec sheets arrive on paper, as PDFs, or as Excel files in each supplier's own format, and staff read them and type the details into the store's product admin. If a page is not updated when a spec sheet is revised, outdated allergen labeling stays up. In April 2026, cashew nuts were added to the allergens that must be labeled in Japan.
The workflow in Kitewell
- Receive spec sheets that suppliers send by email and save the attachmentsEmail
- AI reads ingredients, allergens, origin, and net quantity from the spec sheetAI judgment
- Get what the product page currently shows from the store's product adminWeb
- AI shows, item by item, where the spec sheet and the page differAI judgment
- Staff review the differences and approve the page correctionApproval
- Enter the approved changes in the product admin and save the screenWeb
What AI does
Reads items from spec sheets in different formats and shows where in the sheet each value came from. Allergens are split into those that must be labeled and those whose labeling is recommended. Items read with low confidence go to staff for review.
Where people decide
Staff decide how to correct the product page. AI never changes allergen labeling on its reading alone. When something in a spec sheet is unclear, staff check with the supplier.
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
- How reliably each supplier's spec sheet format can be read
- Which suppliers can send data electronically and which spec sheets need reading
- Whether pages are corrected in the store's product admin or through a multi-channel tool
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
- Could AI change allergen labeling on its reading alone?
- No. AI only shows where the spec sheet and the product page differ. Staff check the relevant part of the spec sheet and approve before the page is corrected.
- Is this needed if we already receive product data electronically?
- Use electronic data exchange for the suppliers that offer it; Japan's Consumer Affairs Agency also recommends it because it causes fewer human errors. This suits the spec sheets that still arrive on paper, as PDFs, or in each supplier's own Excel format.
- 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 product pages across marketplaces
Check prices, comparison prices, origin, and "No. 1" claims on each marketplace and your own site against the product master. Staff decide the fixes.
Check testimonials quoted on product pages
Match reviews and posts quoted on your product pages against your log of commissioned posts. AI flags undisclosed ads; staff decide the fix.
Classify inquiries and draft replies
AI classifies messages from the inquiry form and drafts replies to common questions. A person reviews every reply before it is sent.
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.