Skip to content
Tsuchiya Fuma

Recruiting Strategy × AI Systems

fumawork.com

Hiring, designed.
Kept running with AI.

I extract hiring problems from the business plan and own everything from strategy design through execution. Then I standardize that flow and turn it into AI-powered, automated systems.

Tsuchiya Fuma  — Recruiting strategy consulting × AI systems

For founders and hiring leaders who want recruiting to work

Get in touch See what I do

01 — Approach

Why does hiring keep failing?

Most hiring fails not for lack of strategy, but because upstream and downstream are disconnected: strategy written by people who have never run execution, executed by people who never hear the thinking behind it. That gap hollows out even a good strategy.

I design strategy, KPIs and selection drawing on experience owning the full arc from strategy to execution. The design works backwards from execution, so it survives contact with the field.

And I close that gap with AI and systems, not headcount: design and execution are distilled into structured, automated operations built to keep running reproducibly. That is why a strategy-only engagement still works.

Fig. 1 — The flow * Stages I actually own. Engage me for all of them — or just the design (01–04) or the systemization (06).

  1. 01

    Diagnose

    Problem diagnosis

    Starting from the business plan and mission/vision/values, map out what hiring actually needs to solve.

  2. 02

    Strategy

    Recruiting strategy

    Who to hire, when, and in what order — grounded in research and simulation.

  3. 03

    Metrics & Channels

    KPI & channel design

    Turn strategy into workable numbers and channels.

  4. 04

    Selection Design

    Selection & interview design

    Evaluation criteria in writing, so every interviewer makes the same call.

  5. 05

    Execution

    Hands-on execution

    The designer runs execution first to build the playbook, then shapes it so the team and the systems can keep it running.

  6. 06

    Systemize

    Systemize with AI

    The whole flow gets standardized and turned into an AI‑powered, automated system.

02 — Cases

What actually changed?

* Client names are abstracted to industry and stage per NDA. Every number is a real result.

Case 01

Standing up hiring at a seed‑stage startup

With no dedicated HR in place, owned the recruiting function end to end — from design to execution.

01 Challenge

  • No dedicated HR and no recruiting system in place
  • Needed to hire 10 people in 6 months to fuel growth
  • Leadership had limited time to spend on recruiting

02 Process

  • Diagnose: defined hiring requirements and priorities backward from the business plan
  • Strategy & channels: built the candidate pool across outreach, referrals, and job boards
  • Selection design: documented the interview flow and criteria so anyone could run it

03 Results

  • 30% less leadership time on recruiting via a standardized process
  • Ranked #1 in annual reply rate on a major scouting platform
  • Established a recruiting process that runs regardless of who owns it

Case 02

Hiring for a designer & engineer org

As an engineer-turned-recruiter, led the definition of technical requirements and the articulation of hiring criteria.

01 Challenge

  • Highly specialized, hard-to-fill roles
  • The existing HR team struggled to define the requirements
  • Urgent need for mid-to-senior hires

02 Process

  • Diagnose: used my engineering background to define the technical requirements
  • Selection design: articulated hiring criteria together with the team
  • Execution: built a talent pool and nurtured relationships long-term

03 Results

  • Hired several mid-to-senior engineers and designers
  • Raised awareness through employer branding
  • Transferred specialized hiring know-how to the in-house team

Case 03

Streamlining recruiting ops with AI

Designed and implemented the systemization of the entire recruiting operation.

01 Challenge

  • Recruiting was not standardized and only scaled by adding people
  • Candidate data was scattered, slowing decisions
  • The recruiting team had hit its capacity

02 Process

  • Systemize: automated repetitive work with AI (records, candidate-DB updates, doc generation)
  • Systemize: templatized outreach messages and job descriptions
  • Data design: centralized information so status is visible at any time

03 Results

  • 40% reduction in recruiting workload
  • Decisions can now be made on real data
  • Significantly advanced standardization of recruiting operations

Appendix: Built Systems

Systems built and running at real client teams — what makes the flow repeatable.

Build 01 Candidate ops

Talent-pool foundation (Notion)

A talent pool in Notion that manages candidates and contractors in one place — the foundation every automation below feeds into.

Build 02 Sourcing

Target companies → candidates → scoring & scout drafts

Web-crawls candidates from target companies, attaches evaluations and concerns, and drafts scout messages automatically.

Build 03 Interview records

Interview transcript → summary & DB sync

AI turns an interview transcript into a summary and syncs it straight into the Notion talent pool.

Build 04 Availability

Availability survey → talent-pool sync

A survey asking contractors when and in what areas they can work, mapped field by field into the talent pool.

Build 05 Candidate ops

Candidate auto-update & follow-up reminders

Updates talent-pool fields from interview content and auto-schedules post-interview follow-up reminders.

Build 06 Employer branding

Transcript → employer‑branding article draft

Generates draft employer‑branding articles from interview transcripts.

automation.log illustrative

Build 01 Candidate ops

Talent-pool foundation (Notion)

Challenge
Candidate and contractor data was scattered; every decision meant hunting for it again.
How it works
A Notion database with evaluation axes, pipeline status, and availability as structured fields. Every automation (Builds 02–06) feeds into and reflects back to this base.
Impact
One place for all candidate data — the foundation the other systems run on.
  • Notion
  • Database design

Build 02 Sourcing

Target companies → candidates → scoring & scout drafts

Challenge
Everything from sourcing to first outreach demanded heavy manual effort.
How it works
Define target companies → collect and list matching candidates via web crawling → attach evaluation points and concerns → auto-draft scout messages from each profile.
Impact
Cuts sourcing-to-outreach effort and keeps the talent pool supplied.
  • Web Crawling
  • Claude Code
  • Spreadsheet

Build 03 Interview records

Interview transcript → summary & DB sync

Challenge
Every interview generated note-taking, summarizing, and data-entry chores.
How it works
Claude Code summarizes the transcript and syncs it to the matching candidate in the Notion talent pool.
Impact
Note-taking effort drops sharply; interview data stays current in the DB.
  • Claude Code
  • Notion API

Build 04 Availability

Availability survey → talent-pool sync

Challenge
Hard to know when and in what areas contractors could work.
How it works
A designed availability survey whose answers are mapped field by field into the Notion talent pool.
Impact
Who can take on what — visible with no manual data entry.
  • Google Forms
  • GAS
  • Notion API

Build 05 Candidate ops

Candidate auto-update & follow-up reminders

Challenge
Post-interview updates and follow-ups kept slipping through.
How it works
Interview notes in → Claude Code extracts the fields → talent pool updated via the Notion API → follow-up reminders fired.
Impact
No missed updates or follow-ups.
  • Claude Code
  • Notion API

Build 06 Employer branding

Transcript → employer‑branding article draft

Challenge
Turning interviews into articles took long enough that publishing stalled.
How it works
Transcript in → Claude/GPTs draft → human edits and ships. Generation stops at the draft; publishing stays a human call.
Impact
Shorter lead time to publishing.
  • Claude
  • GPTs

03 — About

Who's behind this?

Tsuchiya Fuma

An engineer-turned-HR professional who thinks across people, organization, and business

I moved from engineering into HR, and now work as a freelancer supporting recruiting for several companies.
Focused on seed to Series B startups, I own recruiting end to end — from strategy design through execution and AI systems.

Approach

Think in structure
Mapping hiring problems from the business plan and mission; connecting frontline execution with leadership decisions
Strategy and execution, both
Executing the strategies I design — so the plan never drifts from reality
AI × HR
Using AI for better, more reproducible decisions — not just efficiency

Career

  1. Dec 2021 – Present

    Freelance HR

    Multiple clients

    • Supporting several companies from recruiting strategy through execution and AI systems
    • Focused on seed to Series B, from kickoff to day-to-day execution
    • Key clients: management-accounting SaaS / design firm / EdTech for kids / travel & hospitality startup / foodtech, and more

  2. Apr 2021 – Dec 2021

    Recruiting HR

    NEXTBEAT Co., Ltd.

    • New-grad and mid-career hiring for engineering, design, and marketing roles
    • 170% of the half-year target achieved
    • Also served concurrently as deputy data-protection officer and department documentation manager, driving systemization

  3. Apr 2019 – Mar 2021

    Full-stack engineer / PM

    NEXTBEAT Co., Ltd.

    • Joined a childcare-tech startup as a new grad and built product as a full-stack engineer
    • Ran new-hire training and mentoring, served as PM for a new business, and handled recruiting in parallel
    • Managed a ~10-person offshore development team in Vietnam

  4. Apr 2018 – Mar 2019

    University years

    Kansai University

    • As a sole proprietor, did door-to-door and in-person sales — ranked #1 nationwide, and managed an 8‑person sales team
    • Ran a career-support NPO (Encourage)
    • During a gap year, traveled to 15 countries, including crossing the Americas and circling India

04 — Services

What can you ask for — and how?

Strategy Design Flow stages 01–04

Strategy Design

Analyze the current state, extract the problems, set strategy and initiatives

  • Leadership & team interviews, problem structuring
  • Talent-market and competitor research
  • Hiring plan, KPI, and channel-strategy design
  • Selection-flow and interview design

Embedded Partner Flow stages 01–06

Embedded Partner

Run the strategy through execution — and keep it running as a system

  • Everything in Strategy Design
  • Execution of the designed strategy
  • An operating playbook your team and systems can keep running
  • AI systemization of the recruiting flow

AI Systems Flow stage 06

AI Systems

Turn your current recruiting flow into AI-powered systems

  • Automated candidate sourcing, scoring, and scout drafting
  • Automated interview records, DB sync, and reminders
  • Built on your existing stack — Notion / Google Workspace / Claude
  • Built so your team can run and adapt it in-house

※ Works standalone, or combined with Strategy Design.

How it works

  1. 01

    Reach out via the form

    A few lines about your situation is enough — "not sure where to start" is fine too.

  2. 02

    Reply & framing

    I review and reply by email, with follow-up questions where needed.

  3. 03

    Proposal & estimate

    Based on the framing, I propose the scope, the way of working, and a cost estimate.

  4. 04

    Contract & start

    We sign a service agreement and the engagement begins.

FAQ

What size and stage of company do you work with?

Mostly seed to Series B startups — I do my best work where there is no dedicated recruiter yet. I also work with companies at other sizes and stages, enterprise included.

What is the contract form?

Independent-contractor service agreements as a rule — monthly retainer or project-based. Details are settled in consultation.

Can you work with clients overseas?

Yes. I work fully remote, and text-based engagements from overseas are welcome.

Can I ask for just the AI build?

Yes. The AI Systems package works standalone — after reviewing your current flow, I systemize the highest-impact stages first.

How do you handle confidentiality?

NDAs are welcome. The cases on this site are published with client names abstracted, under the same obligations.

You don't need a well-formed problem to reach out.

Get in touch

Closing

How do we start?

Reach out anytime. From recruiting strategy through execution to AI systemization — support shaped to your situation.

Fill in the form and I'll get back to you by email. Inquiries from both Japan and overseas are welcome.

* indicates a required field.

Your information will be handled in line with the Privacy Policy.