How to get AI implementation approved in an old-fashioned organization: 'Two practical proposals' from a 57-year-old non-engineer SV fighting on the front lines

On April 7th, something small happened.

An article written by paiza Chairman Ryohei Katayama on note titled 'The Arrival of AI Agents That Sell and Earn: Sierra's Commission-Based Model.' It discusses how we are entering an era where AI agents complete customer interactions and charge based on performance. I wrote a referral article for this piece, titled 'AI Agents Working on Commission and the Japanese Workplaces Where Approval Processes Stall: The Front Lines of a 57-Year-Old SV.'

Gratefully, the author, Mr. Katayama himself, liked and followed my post. The 'field-perspective auxiliary line' I wrote had reached the original author. I was genuinely happy about that.

However, reality doesn't end there.

What was the first thing I did when I arrived at work the next morning?

It was manually stopping the alarm on my smartphone.

A workplace where smartphones ring 10 times a day

I work as an SV (Supervisor) at an outsourced call center. I am 57 years old. I manage about 30 operators and handle about 60 client projects. We operate 24 hours a day, 365 days a year.

The dozens of clients each have their own deadlines. 8:45 AM, 9:00 AM, 10:00 AM, 11:30 AM... Each time, the alarm set on the company-issued smartphone rings incessantly. It rings 10 times a day on weekdays, and twice on weekends and holidays. Whoever notices it stops the alarm and begins the closing process.

This was in April 2026.

A world where AI agents handle 70% of phone calls, and a world where we manually stop smartphone alarms. These two exist on the same timeline. One is a reality within the internet, and the other is the reality right under my feet.

The terminated paid ID

There is another gritty story.

Last fiscal year, after conducting repeated verifications at my own expense, I applied for and introduced a paid Microsoft 365 Copilot ID to the company. In creating QA data for training, I reduced a process that took dozens of hours manually by approximately 71% using Copilot. The ROI exceeded 800%.

However, that ID was terminated as of March 31st.

The contract period with Microsoft was scheduled to end at the close of the fiscal year. I requested an extension in advance, but it was treated as a 're-application' rather than an 'extension.'

No matter how much you achieve, once the fiscal year changes, your authority resets to zero. Anyone who has worked at a large Japanese corporation will understand this feeling. Results are results. But budget is budget. And procedures are procedures. Each runs on a different track.

The two weapons I submitted to the GM today

This is where the main topic begins.

To address these two problems, I wrote a 'proposal' and a 're-application report' respectively, and submitted them to my boss, the GM.

The first is the "Teams Migration Proposal," which replaces smartphone alarm operations with automated Teams notifications.

The second is the "Re-application Report" to continue using Copilot IDs that were cut off after April.

In neither case did I break through with technical prowess. I am not an engineer. I cannot write code.

So, how did I break through?

I "prepared excuses in advance that would make it easy for my boss to stamp their approval."

These two documents are packed not with IT skills, but with the gritty hacks required to get approvals in large Japanese corporations—what I jokingly call "black magic."

In the paid section, I will release these two actual documents along with an explanation of the "intent" I embedded within them.