Back to blog
Manifesto

The Radical Manifesto: AI Recruitment, Done Right

AI is everywhere. The human factor is rare. The Radical manifesto explains who we are, why we built APAC, and why selection is the product.

Nelieke Wismans · Founder of Radical21 August 202611 min read
A person at a laptop, focused on the human work behind the technology

Radical exists because AI changed what hiring needs to find, and the recruitment industry did not change with it. We are a Netherlands-based AI recruitment company. We place AI professionals across Europe on the basis of who they are, not only what they have shipped. We do that with a proprietary framework called APAC, a community we treat as a living room rather than a database, and a tech stack that we own end to end. This is what we believe, and why we believe it.

This piece is the long version of our homepage. If you have eight minutes, read it once. If you only have one minute, the three lines at the bottom are the brand.

AI is everywhere. The human factor is rare.

Models are commodities now. A senior AI engineer can stand up a working RAG pipeline in an afternoon. Fine-tuning a 7B model on a domain dataset is a two-day project for a junior with a tutorial. The frontier moves every quarter, and the cost of entry has collapsed. According to the Stack Overflow 2024 Developer Survey, more than 76 percent of professional developers now use or plan to use AI tools in their work. The technical surface has commoditised. The hard part has not.

What is scarce in the AI economy is not the ability to build. It is the judgment to know when a model is good enough to ship and when it is not. It is the communication skill to defend that call to a board. It is the ethical awareness to refuse a shortcut that looks productive on a roadmap and looks dangerous in production. It is the temperament to stay calm when an eval suite turns red two days before a launch.

Traditional recruitment was built to match keywords to keywords. It works the way a search engine works. Send the query, get the ranked list, fill the seat. That instrument was already a bad fit for software hiring twenty years ago. For AI hiring it is the wrong instrument entirely. A CV cannot tell you how someone handles ambiguity, how they hold a hard conversation across technical and non-technical audiences, or whether they will challenge a shipping decision when the stakes are real. So we built a different instrument.

A person using laptop and phone, the human factor behind every AI workflowA person using laptop and phone, the human factor behind every AI workflow

Selection is the product

Most recruitment agencies sell speed and volume. They optimise for time-to-fill and cost-per-hire. They count CVs sent, interviews scheduled, fees closed. We do not. At Radical, selection is the product. The search, the sourcing, the matching, all of it is in service of one outcome: the people we put in front of you are people we would hire ourselves.

Every Radical clears four gates before they reach a client shortlist.

  1. Applications. Sourced through the network, community referrals, and continuous outbound. We do not do LinkedIn sprays. We do not buy database access. The first thing we look for in an inbound application is whether the candidate read what we wrote.
  2. Interviews. Real human conversations. We look at character, ethics, and ambition. Not keyword overlap. We expect to be surprised, in both directions, often.
  3. APAC test. Our proprietary assessment of the four dimensions that matter most in AI work. No answer key. No leaderboard. The output is a profile that travels with the Radical for the rest of their career with us.
  4. Radical. Accepted into the community. Coaching, peer learning, and events begin long before the candidate meets a client.

Most candidates do not pass. That is the point. Compared to a typical agency funnel, this is the difference we draw.

DimensionTraditional recruitmentRadical
Sold productCVs and time-to-fillSelection itself
SourcingLinkedIn sprays, database buysNetwork, referrals, targeted outbound
Primary signalKeyword match against JDAPAC + human judgment
Candidate pass rateWhoever looks roughly rightMost candidates do not pass
AI useBlack-box screeningAI in the loop, human on the loop
Relationship with candidateTransactional, one placementCommunity, ongoing coaching
Relationship with clientVolume of CVs, racing the clockCurated matches, intake by hand
Compliance posture"We follow GDPR"EU AI Act by default, privacy by design
What we charge candidatesOften zero, sometimes hiddenZero, always. Forever.

APAC: the framework that measures who you are

APAC is our proprietary framework for the human qualities that make great AI professionals. Four dimensions, defined precisely, used consistently across every Radical we accept.

LetterDimensionWhat it measuresWhy it matters in AI work
AAdaptabilityThrives in change. Learns fast. Adopts new tools and contexts quickly. Stays effective as environments evolve.The frontier moves every quarter. A Radical needs to absorb a new framework, a new model family, a new regulatory regime, without losing speed.
PPersonalityAnalytical with a growth mindset. Brings ownership, creativity, curiosity, and ambition to the work.AI work compounds when a team owns its outcomes. We hire for the temperament that takes responsibility before being asked.
AAwarenessUnderstands context beyond the task. Acts with ethical responsibility. Sees the broader impact of decisions.A bad model release is a regulatory and reputational event. We hire for the awareness that prevents it, not the heroics that fix it after.
CConnectionDeeply rooted in personal values. Builds meaningful relationships across teams, stakeholders, and the world beyond.AI products are cross-functional by default. The work fails when the engineer cannot brief the board and the board cannot brief the engineer.

The AI field turns over its tools every few months. A framework that hires for a specific stack ages out within a year. A framework that hires for adaptability, personality, awareness, and connection does not. APAC is how we keep the bar high in an industry where the technical surface keeps moving.

Candidates take the assessment at radicalnetwork.nl. It is free, takes ten to twenty minutes, and lands them in a real human conversation with us. Not a chatbot. Not a recorded interview reviewed by a model. A person.

Human on the loop, always

We use AI in our matching layer. It would be dishonest to pretend otherwise. What we do not do is hand the final decision to a model.

Every shortlist passes through a person. Always. AI is excellent at narrowing a candidate pool from a thousand to ten. It is bad at deciding which of those ten will thrive in a specific team next quarter. That decision needs context no model has: how two founders fight, how a CTO writes Slack, what a CEO actually means by "we want someone senior." That context lives in the intake conversation, the coffee, and the gut feel a recruiter develops over twenty years.

This is also our ethical posture. The EU AI Act, Annex III on high-risk systems in employment classifies AI used for hiring and worker management as high-risk, with binding obligations on transparency, human oversight, and bias evaluation. Compliance is the floor. The principle is higher: responsible AI is the principle that there is always a person accountable for the decision, that the person can explain it, and that the person can be challenged. We will not automate ourselves out of that accountability.

We also publish what the human in the loop does. Every Radical introduction comes with a written rationale from the recruiter who ran the intake. Not a model card. A human one.

A small team in conversation, the human on the loop in every Radical matchA small team in conversation, the human on the loop in every Radical match

European by choice. Sovereign by design.

We host in the European Union. Supabase EU region. Vercel edge in Europe. Email runs through kSuite by Infomaniak, a Swiss provider, on the open IMAP and SMTP standards. GDPR is the floor.

But EU-hosted is not the same thing as sovereign. The transition out of American big-tech reliance is sensible today. It is not a guarantee of tomorrow. A political shift can happen on this continent too. Real sovereignty is the ability to move whenever we want, to wherever we want.

So we built the stack on open standards. Every layer of the Radical infrastructure is replaceable within days. Below is the current and the contingency, side by side, for the four layers that matter most.

LayerWhat we run on todayWhy it is portableWhere we can move next
DatabaseSupabase (PostgreSQL, EU region)Open Postgres, no proprietary extensionsLeaseweb NL, Hetzner DE, self-hosted Postgres
ApplicationNext.js on VercelOpen framework, runs on any Node hostScaleway FR, Hetzner DE, self-hosted Node server
EmailkSuite by Infomaniak (Swiss)IMAP and SMTP, open standardsAny IMAP and SMTP provider, or self-hosted mail server
AI inferenceAnthropic via our matching layerModel-agnostic API abstractionMistral (FR), self-hosted open weights, or any LLM API

Open source went from a hobby project to the foundation of serious enterprise tech in the last two years. Mistral is fully open. Banks and governments are moving to open stacks because they want control over their own infrastructure. We made the same call from day one. A public catalogue of European alternatives exists for every category we use, and we cross-check it whenever a vendor decision needs to be made.

That is sovereignty as a posture, not a slogan. European today, sovereign always.

Love or money

On a business website you rarely read about love. That is exactly what drives us. Love for people and for humanity. Not as a soft concept, but as the foundation for the future. In a world full of AI and automation, human connection is the only guarantee that systems will not drive us further apart.

Money follows when the people part is right. The reverse is not true. That is the order at Radical, and it shows up in concrete choices.

  • The APAC test is free for candidates. Forever.
  • Community access is free for candidates. Forever.
  • Mentorship is free for candidates. Forever.
  • Companies pay us because we deliver real value to companies. Candidates never pay, in any form, ever.
  • We will turn down a placement that pays well if the company is wrong for the Radical.
  • We will turn down a Radical that scored high on APAC if their character did not match the report.

This is what "love or money" means in operations, not in marketing. If a decision pulls us toward money at the cost of the people part, we do not take the decision.

What this means for the people we work with

For candidates. A real human reads your application. We will not introduce you to a company we would not want our friends to work for. You stop being a commodity the moment you are accepted. You join a community of doers, thinkers, and builders with a personal mentor, peer-to-peer learning, quarterly check-ins, and access to roles you will not see on any job board. We tell you the truth about your trajectory, including the parts that hurt. We are direct because we owe you that.

For companies. We will tell you the truth about your role, your timeline, and your fit, including when it is uncomfortable. We will not send you a CV until we believe it should land on your desk. Every Radical on your shortlist has cleared four gates including APAC. You are hiring for judgment, not only for a CV. We respond within twenty-four hours. We stay in touch long after the hire, because a placement is the start of a relationship, not the end of a transaction.

We do not promise speed. We promise that the person we introduce is someone we would hire ourselves.

Amsterdam canal houses, where Radical is based and where the European-by-choice principle startsAmsterdam canal houses, where Radical is based and where the European-by-choice principle starts

The team behind the manifesto

Radical is founder-led by two people with more than 45 years of combined experience in identifying talent.

Nelieke Wismans spent over twenty years in executive search before founding Radical in 2024. Her mission is to guide AI-natives in balancing rational thinking with emotional awareness. She believes the qualities most critical for responsible AI are empathy, ethical awareness, intuition, and contextual judgment. They are the foundation of our APAC selection model, and they are why she actively recruits and advocates for women in AI.

Oscar Voskuil has spent over twenty-five years building commercial engines for companies at every stage of growth, from early startups to organisations in transformation. His conviction is consistent across that span: business only works when it works for people. He built Radical to change how talent is identified, assessed, and placed in an industry where AI is reshaping everything.

A small core team runs the day-to-day. Vincent Zepeda, Head of Technology, turns recruitment instinct into intelligent systems. Every layer of the Radical stack is his. Alongside that, a small team builds the spaces where AI-natives grow, learn, and find their people.

We are accountable. The founders are visible. The names on the about page are the names you will meet.

What we will not become

A boundary statement, written down so we can be held to it.

  • We will not become a CV mill, no matter how much revenue volume promises.
  • We will not become a LinkedIn sprayer, no matter how cheap the inbound becomes.
  • We will not become a black-box AI matching service that hides behind an algorithm.
  • We will not become a faceless agency where the founders are invisible and the recruiters change every six months.
  • We will not become a company that puts speed above selection.
  • We will not become a company that locks itself into a single cloud, a single model, or a single jurisdiction.
  • We will not become a community in name only, with no real events, no real coaching, no real care.
  • We will not become a US-default tech stack with European marketing.
  • We will not become a consultancy that charges by the hour instead of standing behind the hire.

When a decision moves us toward any of those, we stop and ask a different question.

The future we are building

We are building the European standard for AI hiring. A standard in which the human factor is the first criterion, not the last. A standard in which every match comes with a name, a face, and an accountable human behind it. A standard in which a community of vetted AI professionals develops together, holds each other to a higher standard, and shapes the foundation for ethical AI on the continent.

We do this from Amsterdam, with a small team that runs the community, the technology, and the growth. We do this for AI engineers, researchers, MLOps people, AI product leads, and AI architects who want to be selected for who they are, and for companies that hire them on the same standard.

Human on the loop. European by choice. Trustworthy by design.

That is the future we are building. Welcome to Radical.

Frequently asked questions

The Radical manifesto sets out why Radical exists, how we select AI professionals, and what we refuse to compromise on. It is built on three principles: a human on the loop in every matching decision, European by choice on infrastructure and ethics, and trustworthy by design through portability and open standards.

Sources

  1. EU AI Act, Annex III: AI used for recruitment and worker management is classified as high-risk (point 4)artificialintelligenceact.eu
  2. Stack Overflow 2024 Developer Survey: AI section, 76% of respondents use or plan to use AI toolssurvey.stackoverflow.co
  3. Mistral AI: open-weight model releases on Hugging Facehuggingface.co
  4. GitHub Octoverse 2024: the state of open source and AI on GitHuboctoverse.github.com
  5. European Commission: Open Source Software Strategy 2020 to 2023commission.europa.eu
  6. European Alternatives: catalogue of European replacements for popular SaaS and cloud serviceseuropean-alternatives.eu
Looking for AI talent?

Tell us what you need.

We respond within 24 hours, from a real human.

Get in touch