This week on the EUVC Podcast, Andreas Munk Holm sits down with Matthew Wilson, co-founder of Jack & Jill, and Peter Specht, General Partner at Creandum. Fresh off a $20M seed to take their AI recruiting agents global, they dig into how conviction is built in Europe, from founding insight to investor belief, and what it now takes to scale an agent-native company with speed, precision, and craft.
Jack helps candidates find and optimize their careers. Jill helps companies hire brilliantly. Together, the two agents form a high-signal, two-sided network that aims to become the world’s most networked AI-powered recruitment agency — without the classical incentive conflicts of human middlemen.
🎧 Here’s what’s covered:
02:35 | Why Creandum leaned in, conviction on voice-based interfaces and why recruiting is a massive, broken vertical for agent AI
03:38 | The founding moment: leaving Omnipresent, 18 months in the wilderness, and the February insight that agents make talent marketplaces finally viable
07:07 | Recruiting is broken (and AI made it worse): why first-principles thinking is needed to avoid “more noise, not more signal.”
09:15 | Investor conviction: founder/market fit, why this moment is different, and the defensibility of a two-sided agentic marketplace
12:22 | The user experience: the “coffee chat” with an AI recruiter: deep voice conversation → matching, prep, coaching, introductions
16:30 | Solving the incentives trap: why Jack works 100% for candidates and Jill works 100% for companies (fixing agency conflicts)
19:10 | Coaching as core: how AI unlocks career guidance, interview prep, and hands-on support that humans rarely get today
22:47 | Building fast in the AI era: talent density, global expansion, and why a 20M seed makes sense for a dual-product marketplace
26:35 | Two companies in one: scaling Jack (consumer) + Jill (B2B) simultaneously, across markets, with AI leverage
34:02 | The GTM playbook: engineering-led marketing, AI-driven creative testing, instant value, and rethinking B2B buying entirely
37:47 | The new AI go-to-market: speed, PLG dominance, virality-by-design, and why distribution now matters more than ever
43:52 | Two GTM worlds: viral AI products vs. slow, enterprise-heavy AI deployments (and why both will coexist)
47:15 | The “productization” of marketing — why engineering now powers growth, not headcount-heavy marketing orgs
50:29 | Final advice (VC POV) — start with a unique insight, not a trend; think in 5–10 year arcs, not quick ARR bumps
🎧 Listen on Apple Podcasts and Spotify 🎧
🎯 This Episode’s Themes
Two agents, clean incentives: Jack represents candidates 100%; Jill represents companies 100%. No conflicted middleman.
Voice + agency > automation: Deep conversational intake to understand people and roles — then agents that actually work on your behalf.
Networked marketplace with memory: Every conversation compounds a proprietary, high-signal network across candidates and companies.
Why now (and why $20M): A massive, broken market, clear “why now” for agentic AI — and rare execution speed that justifies an outlier seed.
Talent density as a weapon: AI amplifies top performers; small, ultra-strong teams beat large orgs built the old way.
AI-native GTM: Product-led growth, virality, and AI-powered marketing + sales motions that break the classic SDR/AE bottleneck.
Building for 2030, not 2025: Don’t just automate today’s workflow — design for the world with 2–3 more model generations.
✍️ Show Notes
Founding story & why now
Matthew’s path to Jack & Jill runs through Omnipresent, his previous startup, which built infrastructure for companies to hire globally and helped tens of thousands of people access opportunities worldwide.
After handing over the reins in 2023, he spent 18 months exploring what to do next, with three hard criteria:
Mission: A problem he deeply cares about, improving people’s lives at scale.
Ambition: A concept with genuine potential to become a globally leading, category-defining company.
Co-founder fit: Someone brilliant with aligned values and complementary skills.
In February–March, conversations with co-founder Saaras around agentic AI and recruiting crystallized the thesis:
Recruiting is a huge, broken market.
AI had reached a point where agentic products (not just autocomplete) could act on behalf of users.
Voice and better matching could finally make a marketplace that isn’t just “LinkedIn with more noise”.
They spent roughly a month stress-testing: Why did past talent marketplaces stall? What’s actually different now?
The answer: agent AI, voice, and the ability to truly work for each side — not just push more candidates into inboxes.
What Jack & Jill actually do (and how it feels to use them)
For candidates (Jack):
You share your LinkedIn (and maybe a CV).
Then you jump into a voice conversation with Jack — the part Matthew describes as the “beer or coffee chat” you’d have with a trusted friend.
Jack doesn’t just ask for title, salary, and location. He explores:
what you actually enjoy doing,
what you don’t want to do again,
your views on certain markets,
culture and team preferences,
what “ideal next step” really means for you.
Once he has that deep context, Jack:
Scans every public job listing he can find and surfaces relevant ones in real time.
Helps refine applications (CV feedback, cover letters, role-specific tailoring).
Runs mock interviews for roles you’re targeting.
Facilitates direct intros into companies where Jill has done similar deep discovery on the other side.
Interfaces where you already are: email, WhatsApp, with a web interface if you want it — but it’s not required.
For companies (Jill):
You brief Jill like you would a human recruiter:
what success looks like,
who thrives inside your culture,
what “must haves” and “nice to haves” really mean, beyond the JD.
Jill builds a deep profile of the company and open roles, then:
mines the candidate pool Jack has spoken to,
surfaces and introduces high-fit candidates,
supports more of the recruitment workflow over time.
Two agents, one marketplace — and no conflicted middleman
Human recruitment agencies sit in the middle and get paid when a hire closes. That creates incentive misalignment:
They’re selling both sides.
They’re optimising for their fee, not the long-term fit.
Jack & Jill flip this:
Jack works only for the candidate and holds information they’d never share directly with a company.
Jill works only for the company and collects information that candidates wouldn’t see.
Jack optimizes for the candidate’s long-term benefit; Jill optimizes for the company’s hiring success.
“We trust that if Jack truly works for the candidate and Jill truly works for the company, we’ll build the best product for both — and the business will grow as a consequence.”
On top of this, AI is ideally suited to keep perfect memory of every data point over years across both sides, creating:
better fit than a human could reasonably manage,
a compounding moat of proprietary, structured preference data.
The $20M seed round: why it’s an outlier and why it made sense
Peter is clear: most European seeds are still in the €2–8M range.
Jack & Jill’s $20M seed is very much an exception, and for Creandum it rests on three pillars:
Founders:
Matthew (Omnipresent) and Saaras (AI SDR tooling) both come from fast-paced, sales-heavy categories (EOR and AI sales) with similar target customers.
They’ve done company building at scale before and know how to deploy capital.
Market & product:
Recruiting is one of the largest, most painful markets for both companies and individuals.
Jack & Jill are not “automation on top” — they’re a two-sided, agentic marketplace with network effects and defensibility.
Execution so far:
In the first 6–8 months, the team reached tens of thousands of candidates, with strong engagement and user feedback.
For Creandum, that early velocity supports the case for backing them at “category leader” scale.
The round is sized so that Jack & Jill can:
build out both consumer and B2B products properly,
invest in seeding multiple markets early (London, San Francisco, more),
and capitalize the company to move at the pace this category now demands.
Strategy: two businesses in one – and a smaller, sharper team
Matthew openly says they’re building two businesses inside one:
A consumer product (Jack) with its own growth, support, and UX.
A B2B product (Jill) with its own sales, onboarding, and lifecycle.
On top of that, they’re going multi-market from very early:
Already active in London.
Mid-launch in San Francisco.
Planning further geos to truly own “top-of-funnel globally” on both sides.
Rather than scaling headcount like Omnipresent (2 → 450 people in three years), Jack & Jill are built around talent density:
AI massively amplifies top performers (Matthew estimates 10× performers could become 20–40× with AI).
That means you don’t grow by adding hundreds of ordinary people, you find 30–40 exceptional people and build around them.
“It’s much more enjoyable to run a small, talent-dense company than a 400–person org — and you can out-execute bigger incumbents.”
AI-native go-to-market: consumer + B2B
Consumer / candidate side (Jack):
Heavy use of AI for content creation, creative testing, and performance marketing.
Ability to iterate through huge volumes of creative, target segments, and messages at very low marginal cost.
A tight loop:
creative → distribution → who activates → predict user value → feed back into targeting and spend.
B2B / company side (Jill):
Matthew is unapologetically anti-legacy in B2B sales:
The classic pipeline (SDR → AE → AM → CSM) is slow and painful for buyers.
Buyers who want to purchase wait weeks to be “qualified” and “taken through a process”.
Jack & Jill’s ambition:
Use agentic AI to answer questions dynamically, in real time.
Have Jill act as the best SDR + AE + CSM rolled into one, across the whole customer lifecycle.
Break the constraint of “we grow at the pace we can hire and ramp AEs”.
Peter adds the broader market context:
The fastest-growing AI companies so far (Lovable, 11Labs, etc.) have been PLG-first: magical product → virality → enterprise later.
Product-led growth, social visibility, and founder storytelling are now core GTM levers, not side activities.
At the same time, there is still a heavy-enterprise wave of AI companies with classic, relationship-driven motions — it’s not one-size-fits-all.
Productizing marketing & ops
Andreas floats the idea that marketing is being “productized” — instead of five marketers + an agency, you build internal AI “machines” that:
read customer feedback,
generate content,
launch and iterate campaigns automatically.
Matthew agrees in spirit and reframes it as engineering embedded in every function:
They’re building the engineering team into marketing, sales, and ops, not just the core product.
“Builders” and toolsmiths inside each function design the systems that run far more work than a human team could.
Peter’s view:
This is becoming the new normal for high-performing AI companies — more in-house capability, fewer outsourced processes, and a new skill set around orchestrating tools and agents.
Closing reflections for founders & VCs
Matthew’s two big points for founders:
Don’t be incremental.
Too many AI startups are just automating an existing workflow. The defensible ones project:What will this market look like in 2028–2030 with 2–3 more generations of models?
Then build the product and company for that world, not today’s.
Hire differently.
Talent density matters more than ever.Small, brilliant teams, amplified by AI, can out-execute large incumbents.
Design your org for leverage, not vanity headcount.
Peter’s lens for investors and founders thinking like investors:
Always ask: what’s your unique insight?
Are you just following the market?
Or do you see something about how the market evolves over 5–10 years that others don’t?
The game isn’t getting to a few hundred K ARR; it’s building something that can be truly valuable in the long run.
💡 One-Liner Takeaway
Jack & Jill show what an AI-native marketplace looks like when you start from incentives, memory, and GTM — not just automation. If you’re thinking about agentic AI, recruiting, or the next generation of PLG companies, this is one to study closely.
