AI implementation for service businesses

AI around the CRM you already have

I am Kotaro Izumida, an engineer based in Japan. I connect AI to the CRM and tools a business already uses, so the people stuck on office work can get back to the work that grows it. A person approves what goes out.

What I build

Drafts that a person approves

AI drafts quotes, replies, and summaries from your own records. A named person checks each one before anything reaches a customer.

AI around the CRM you keep

Your CRM stays the place where records live. I add AI and automation around it instead of replacing it.

Keeping it running

Tests, release steps, access rights, and handover notes are part of the job, not extras for later.

Track record

Zoho Developer Community Hackathon 2025

Listed among the Category Winners

AI Interview System, a first-round hiring interview flow built with Zoho SalesIQ, CRM, Creator, Bookings, and OpenAI.

See the official winners page

US home services company (name withheld)

Ongoing Zoho CRM work

  • Deal stages decided by rules, not by each person's judgment
  • Stage overrides limited to named managers
  • Credit card fee toggled with one click
  • Owners assigned automatically by ZIP code

More than three years of building and customizing Zoho, and implementing with AI. Client reviews and job history are on my Upwork profile.

How we work

  1. 1Tell me about one back-office task: what it is, who does it today, and how often it comes up.
  2. 2We measure it and agree on a small test, so the result can be compared with how it works now.
  3. 3I build the draft and the review screen first. Nothing goes out without approval.
  4. 4After the test, you decide to keep it, change it, or stop.

I am in Japan, so the work moves through written messages, recorded walkthroughs, and async updates. Every decision and next step stays in writing, so nothing depends on catching me live.