AI that grows revenue without adding headcount.

ChallengeQuotes built up from material and labor costs according to the specification depend on one particular person. Responses take days, and stop when that person is away.

1. InputQuote requests including drawings and specifications
2. AIOrganizes materials, quantities, processing and other details from the request, and searches for similar past jobs
3. CalculateCalculates the estimate from cost data and pricing rules, and produces a draft quote showing its basis
4. ApproveStaff set the final price

ExceptionsDrawings that are hard to interpret, and specifications with no precedent, are estimated by staff

Metrics

Days to quote, quotes per person, win rate

Human role

Deciding and approving the final price

Key considerations

Organizing past quotes and cost data; writing down unwritten rules such as discounts

ChallengeDeal notes and follow-ups vary from one salesperson to the next, and proposals take a long time to write. When things get busy, follow-ups slip and deals are lost.

1. InputCall recordings, meeting notes and customer emails
2. AISummarizes key points, the customer’s needs and next steps
3. AIDrafts the proposal and follow-up email from customer data and past proposals
4. SyncLogs the meeting and tasks in your CRM or kintone

ExceptionsAnything involving price or contract terms is checked by the salesperson before it goes out

Metrics

Time to prepare proposals, days to follow up, win rate

Human role

Deciding on the proposal and terms, building customer relationships

Key considerations

Rules for handling recordings, tidying up CRM fields, standard proposal templates

ChallengeProduct specifications and past cases are scattered across shared folders and the heads of a few experienced employees. Answers take time, and quality depends on who responds.

1. InputTechnical inquiries from customers
2. AISearches product documents and past support history for relevant information
3. AIDrafts an answer, citing the supporting documents
4. ReviewStaff review the draft and reply

ExceptionsQuestions with no support in the documents go to staff without a draft

Metrics

Time to answer, variation in answer quality between staff

Human role

Checking answers; keeping documents and knowledge up to date

Key considerations

Organizing documents and keeping them current; limiting what can be referenced according to access rights; designing it to avoid unsupported answers

ChallengeOrder forms arrive from each customer in a different format (fax, PDF, email) and are keyed into the order management system by hand. In busy periods, overtime and input errors rise.

1. InputOrder forms arriving by fax, PDF or email
2. AIHandles format differences and reads item numbers, quantities and delivery dates
3. AIChecks against customer and product master data, converting customer-specific codes and spelling variations to your own item numbers
4. SyncEnters the order into the order management system (via API, CSV import, etc.)

ExceptionsItem numbers missing from the master data or price mismatches are checked by staff before registration

Metrics

Entry time per order, input errors, overtime in busy periods

Human role

Reviewing and approving exceptions

Key considerations

Handling varied formats and spellings, poor fax quality and handwritten corrections; how to integrate with the order management system

ChallengeLarge volumes of invoices arrive every month, and are matched against purchase orders and entered into the accounting software by hand. Japan’s Qualified Invoice System (consumption tax invoice rules) and Electronic Bookkeeping Act have added to the checking workload.

1. InputInvoices arriving as PDFs or on paper
2. AIReads amounts, consumption tax for each tax rate, and invoice registration numbers
3. MatchVerifies registration numbers on the National Tax Agency’s public site, and matches against purchase order and goods receipt data
4. SyncRegisters draft journal entries in the accounting software (AI suggests the accounts)

ExceptionsAmount mismatches, or registration numbers that can’t be verified, go to accounting staff for a decision

Metrics

Days to close the month, hours spent matching and entering

Human role

Reviewing draft entries and mismatches; month-end close

Key considerations

Matching rules for purchase order and goods receipt data; integration with the accounting software; designing storage to meet Electronic Bookkeeping Act requirements (to be confirmed with your tax accountant)