What the inside sales floor taught me is that the will to finish decides results more than strategy does. However good a system is, it means nothing if it doesn't run on the ground.
Generative AI is the same. Large enterprises already use it well, and the gap keeps widening. We go into small and mid-sized companies and stay with them until the technology is genuinely usable. Making Japanese sales strong again — that is what drives our next step.
Formerly at CyberAgent. After leading marketing support for the world's largest air-conditioning manufacturer, he founded Wabisabi in 2023.
This page walks through each of the three in detail.
The sales floor of fast-growing SaaS companies, and generative AI rollouts at small and mid-sized firms. Even junior members talk directly with decision-makers and work from first-hand information.
The Kizashi AI series is built in-house. In sales you are its first daily user; in engineering you grow the product with data that carries real outcome labels.
What we look at, how you move up, and what it pays. Our grades, reviews and compensation logic are open to everyone. Seniority and tenure play no part.
Winning sales patterns honed on the floor, plus generative AI. Two businesses to choose from, and the in-house products behind them.
Keep SaaS pipelines moving
Target design and list building, appointment setting and qualification, script improvement. We work as part of the client's own team and commit to results.
The finish line is results, not a training day
Generative AI training that raises win rates, and Forward Deployed Engineers who embed it on site. We give smaller companies weapons that hold up against the majors.
Founded in Umeda, Osaka. Inside sales support specialized in SaaS / BpaaS begins.
Creating pipeline for fast-growing SaaS companies in business cards, expenses and HR.
Cumulative meetings pass 8,000 across 60+ client companies. The inside sales playbook is codified.
Generative AI training and FDE launch. The in-house Kizashi AI series goes into full deployment.
Plenty of companies do sales support. Plenty build AI products. Because we do both under one roof, what we learn at the exit — the closed deal — flows back to the entrance: who to call, what to say, and how.
5.82M companies observed monthly; the ones that move are spotted early
▶Messages tested under equal conditions; only the winners go out
▶Calls are analyzed and the right line appears on screen within two seconds
▶Drafted from the audio with evidence attached; flawed appointments are stopped before hand-off
▶Sales-specific generative AI training that leaves capability on the floor
A company's movement only becomes a signal once monthly records stack up. You can only build differences from the month you started observing.
Which company, when, who spoke, what was said and how it came back. We can trace how each company's thinking developed.
Which script won the meeting, and which meeting reached a signed deal. That mapping can only be built by someone who makes the calls and follows them to the close.
You become the first daily user of AI we build ourselves, and what you ask for on the floor goes straight into the next release.
You grow the product with first-hand field information and outcome-labelled data. The people who use it sit at the next desk.
You deliver a curriculum born on our own client floors, in your own words. The source material is our real internal data.
What we look at, how you move up, and what it pays. Our grades, reviews and compensation logic are open to everyone. Decisions rest on numbers and records, not seniority or tenure.
The formula and the rates are open to everyone. Pay is decided exactly as written here.
It does not vary by grade and never goes down (fixed overtime included). We share the figure during the hiring process.
Paid while you are responsible for people as a leader or manager. It pays for the role you hold now, not for your ability.
Tied to the gross profit your team generates. Calculated from zero every month, never carried over.
Paid in any month where gross profit reaches ¥300K or more. Revenue from leads qualified by part-time staff also counts as yours.
| 9% | Gross profit ¥300K – under ¥1.5M |
| 10% | Gross profit ¥1.5M – under ¥2M |
| 13% | Gross profit ¥2M and above |
Promotion changes this add-on. Base salary does not change with grade.
You own your numbers, and by the end you can hand your playbook to someone else.
You create results through people, until the team runs without you.
Responsible for the profit of a department, a business, or the whole company. This career sits beyond the leader track.
These figures model a standard level of performance at each grade. The incentive tracks monthly gross profit and has no cap, so results can take you above them.
The figures combine base salary, fixed overtime, role allowance and an assumed incentive. Commuting costs are reimbursed in full on top. Your starting grade is decided individually, based on experience and the interviews.
The revenue set for your grade, three months running. Judged automatically from system data — no one's opinion enters into it.
Nine job criteria — communication, management, generative AI, problem solving and more — at the level your grade requires. Judged from monthly interview records and performance data.
Our ten values, measured by the actions you actually chose rather than your agreement with the words. Anything that breaks trust halts the review immediately.
One concrete example is required. What counts as having done it is defined specifically for each grade (see below).
There is no essay to write and no paperwork to file. You answer a three-minute bot interview at month end. What you say is kept as a record, and you can read your own record any time.
The AI only organizes records; it makes no judgement. Promotions are decided in a meeting with the CEO, comparing records against performance. You are told the outcome and the reasoning on the spot.
The same criterion is held to a different standard as you move up a track. We do not ask leader-track standards of a G1.
| Criterion | Player track (G0–G3) | Leader track (C1–C4) |
|---|---|---|
| Communication | States facts clearly and shares them before being asked | Resolves misalignment early, inside and outside the team, and builds agreement |
| Leadership | Influences those around them through their own actions | Sets a goal, moves people, and produces results |
| Management | Manages their own tasks and deadlines | Manages the team's numbers, shifts and progress |
| Teamwork | Helps colleagues and asks for help when needed | Designs how cooperation works inside the team |
| Adaptability | Tries new methods and learns them quickly | Translates change for the team and makes it stick |
| Reporting and escalation | The worse the news, the faster it is passed on | Embeds the reporting habit across the team |
| Learning | Learns the product and the industry on their own | Turns learning into shared knowledge for the team |
| Generative AI | Uses AI fluently in daily work | Creates the team's pattern for using AI |
| Problem solving | Solves the problem in front of them | Solves team-level problems with systems |
Judgement never rests on self-assessment. Each criterion is marked met or short using verifiable facts: monthly interview records, system performance data, and day-to-day work. Where something falls short, you are told what is missing and what to change over the next three months.
Carried a taught procedure through to the end without asking
Acted on their own in a down month and brought the numbers back
Taught their method to someone else, and that person's numbers moved
Taught several people, and it ran without them
Filled a gap in the team with a system
The chart shows the proof required to move up a grade. On top of that, you must meet the revenue set for your grade three months running (we share those figures during the hiring process). From C1 onward, a leader's results are the team's numbers and the people who grew there.
Not stiff, but seriously committed. We enjoy the obstacles and win as a team.
ARCA Umeda Building 11F / 5F, 4-3-32 Nakazakinishi, Kita-ku, Osaka. Walking distance from Umeda Station.
One employee plus two part-time members is the base unit, with daily coaching, weekly 1-on-1s, morning huddles and knowledge sessions.
Generative AI is one of the nine job criteria. Every employee uses generative AI and AI agents in their own work as a matter of course.
Fast-growing SaaS companies and live AI deployments. Even junior members talk with decision-makers and turn winning patterns into playbooks themselves.
In any month a team hits its goal, that team goes out together at the company's expense. There is a monthly office gathering as well.
The pay formula, the grade requirements and the review process are documented and given to everyone. Nothing comes as a surprise after you join.
This is the profile behind both hiring and reviews. The six core values apply to everyone; the decision-making and leadership principles apply from the leader track (C1 and above).
Faces facts and changes their thinking or behaviour when it is warranted.
Never be ashamed to ask / Never hesitate to correct a mistake
Chooses honest action for clients, colleagues and the company. Never uses ability or results as an excuse for arrogance.
Without trust nothing stands / No brilliant jerks
Owns what was produced, not how hard they tried.
The burden is heavy and the road is long
Gives those around them hope and energy, and keeps asking what to change next.
Understands the actual work and the actual numbers, and builds systems that deliver alongside the people doing it.
Work for the happiness of others
Is not ruled by emotion, and chooses words and conduct that fit the situation.
Too much is as bad as too little
Argues candidly before a decision, and owns the success of the direction once it is made.
Harmony is to be valued / Have backbone; disagree and commit
Puts what they learn to work and gives it back in a form others can repeat.
Brings the right people together to solve problems outside their own remit. A leader's results are the team's numbers and the people who grew there.
All roles are based in Umeda, Osaka, and are full-time. You can start with a casual conversation.
Come and talk to us first. You can start with a casual conversation.
Ask us anything about the business, the policies or the team. This is not part of the selection.
We talk through your experience and what you want to do at Wabisabi.
A direct conversation with Shota Yagi, covering the values and agreeing your starting grade and terms.
We put the terms in writing, and agree your start date together.
At G0 the gate is simply hitting your activity target and closing your first deal. We do not ask for high revenue from day one. Scripts, recordings of senior colleagues' meetings, role-play, and the live rebuttal prompts in Kizashi AI Dialogue — the playbook is all there. And in a month without results, base salary is still paid in full.
The numbers are judged automatically by the system, so impressions do not move them. Work and values are judged only on verifiable facts, and promotion is decided in a meeting with the CEO. Numbers and records get you promoted, not someone's opinion of you.
Numbers alone never cause a demotion, and base salary is designed not to fall. Where something falls short, we tell you specifically what is missing and what to change over the next three months. In April, August, November and December — when the market is slow — targets are judged at 80% of the standard figure.
Individually, based on your experience and the interviews. Seniority and tenure play no part. Base salary is the same at every grade, with the role allowance and the monthly incentive stacking on top. The annual figures by grade are on this page, and the terms are given to you in writing with the offer.
It depends on the role. FDE requires experience building and deploying with LLMs and cloud AI. Inside Sales does not require prior experience, but once you join, everyone uses generative AI daily — it is one of the nine job criteria. The appetite to master it is needed in every role.
Yes. The casual conversation in step 1 is not part of the selection. It is there for you to ask about the business, the policies and what the team is like. You do not have to decide anything on the day.
Casual interviews welcome. Apply or ask us anything from here — we can also share the full documentation on our policies.
Apply or get in touch