pre-seed · raising now · EU + US

The delivery network for agentic hiring.

AI agents are entering recruiting. They need shared rails to negotiate, match, and settle — compliantly, globally. ADNX is that infrastructure.

The problem
The opportunity
ADNX
$4,700
avg cost per hire (SHRM 2024, USD)
$620B
global staffing market (SIA 2025)
40+
structured signals per side
5+ months
avg job search duration (Randstad 2024)
50M+
workers change employers yearly (US + EU)
<5 min
agent-to-agent match cycle
~80%
of agency workflow still manual (est.)
100K+
recruiting agencies globally
1
integration, full network access
3.7%
application-to-interview rate (Huntr 2025)
0
compliant agentic hiring rails
2%
infra take rate on every hire

Problem → how we solve it

Recruiting plumbing is broken. We're building new rails.

No shared standards, black-box matching, compliance minefields, bespoke integrations. ADNX fixes each with open protocols, bilateral agent negotiation, built-in compliance, and one-integration network access.

No shared standards

Every ATS and job board uses different schemas. Agents can't interoperate.

Open protocols (OTP + OJP)

MIT-licensed specs for talent and jobs. 40+ structured signals per side. Machine-readable.

Black-box matching

Centralized keyword scoring. No bilateral negotiation. No audit trails.

Bilateral agent negotiation

Each side has an autonomous agent with hard constraints and soft preferences. Agents iterate until match or pass.

Compliance minefield

EU AI Act (high-risk, €35M fines), NYC Local Law 144, Illinois AIPA, Colorado AI Act. Every market is regulating AI hiring. No one has built compliant infra.

Compliant by architecture

Immutable audit trails, human-override logs, compliance vault. Works across jurisdictions — EU, US, global. Built in, not bolted on.

Point-to-point doesn't scale

Every agent integration is bespoke. This is the pre-Stripe era of recruiting.

One integration, full network

Integrate once — agents discover, match, and settle with any node. Network effects compound.

Agents build your pipeline in real time.

Employer Agent OJP
role Staff Engineer
stack Ruby, Go
budget €90–110K
location Berlin
↔ adnx
Talent Agents OTP
profiles 312 on network
protocols OTP structured
signals 40+ per candidate
negotiate bilateral, autonomous
Staff Engineer
Berlin · Ruby, Go · €90–110K
scanning
25 candidates evaluated. Agents negotiate fit, salary, and terms bilaterally — top matches surface automatically. No recruiter. No job board.

Model & traction

Infrastructure economics. Already building.

Revenue streams

ADNX Hire — lightweight hiring interface for SMBs. Not another ATS — agents do the sourcing, screening, and matching via the network. The employer just reviews and decides. Tiered freemium model. 57M SMBs in US + EU (SBA, Eurostat) — vast majority lack dedicated hiring tools. Every signup = a new node.

Network fees — percentage of first-year salary on every hire settled through ADNX. Small at first, dominant at scale. This is the Stripe model.

Agency white-label — agencies run their own agents on ADNX rails. Per-seat pricing.

Traction

Private beta — ADNX network live, bilateral negotiation working

Protocols published — OTP + OJP, MIT licensed, A2A-compatible

EU AI Act ready — compliant by design

Market sizing (base case)

Layer
Size
Scope
Global staffing
$620B
Contextual. ADNX does not address all of this.
Network fee pool (EU + US)
~$1.55B
2% take-rate on agency-mediated placements
SaaS opportunity (EU)
$100–250M
2–5% of EU ATS market via freemium + agentic capabilities
Y3 target (base case)
€14M
Bottoms-up build. Details in full memo.

Revenue model details, unit economics, and financial projections available in the full IC memo upon request.

ADNX Hire (SMB) Network fees Agency white-label seats

Why now

AI agents are entering hiring — they need shared rails, not more point-to-point APIs

Regulation is accelerating globally — EU AI Act high-risk enforcement from 2026 (Omnibus may extend to late 2027, but backstop deadlines hold). NYC LL144 already active. Illinois AIPA, Colorado AI Act imminent. Every market needs compliant rails — the direction is clear regardless of timing.

No compliant rail exists — first mover sets the standard. Compliance is the moat.

The round

€2.5M Post-Money SAFE · €12.5M cap · 20% discount · MFN

20% dilution (fully diluted, incl. 10% ESOP reserved) · Pro-rata at €100K+

~24 months runway — Berlin-based, EU launch, DACH-concentrated sales

40% GTM & customer acquisition
30% engineering
15% legal & compliance
15% ops

Unlock full investor brief

Enter your work email to see financials, round details, and the full value creation plan.

Investment thesis

Three core arguments.

01

Regulation creates mandatory infrastructure demand

EU AI Act classifies AI hiring as high-risk (Annex III). Original high-risk enforcement: August 2026. The EU Digital Omnibus package (proposed Feb 2025) may extend this to late 2027 — but backstop deadlines ensure enforcement regardless, and US regulations (NYC LL144, Illinois AIPA, Colorado AI Act) proceed independently. Even under delay, SaaS revenue is deadline-agnostic and early adopters gain advantage. This is not a discretionary purchase — it's a regulatory mandate on a known trajectory.

02

Network effects from open infrastructure

Open protocols (MIT-licensed) maximize adoption. Monetization happens at the proprietary execution layer: negotiation engine, compliance vault, settlement. Each SMB on ADNX Hire and each agency on white-label adds a node. Unlike LinkedIn's walled garden, ADNX is neutral infrastructure.

03

Distribution wedge via SMB ATS

ADNX Hire is not a standalone SaaS business — it's a distribution mechanism. Every SMB onboarded becomes a network node. Free tier generates liquidity even without conversion. The "customer" is actually inventory for the network. This inverts the typical SaaS acquisition cost model.

Competitive landscape

No one has built this layer.

Enterprise ATS (Personio, Workday) are walled gardens. Agentic tools (LinkedIn, SeekOut) are application-layer. SMB HR-tech lacks agent capabilities. AI RegTech is horizontal. None offer open infrastructure with matching, compliance, and settlement.

Enterprise ATS

Personio, Workday, SAP, Greenhouse. Walled gardens. No agent interop. Could ship a compliance module — but that's a feature inside their garden, not a network. ADNX integrates with Personio; at scale, it becomes the neutral layer Personio's competitors also connect to.

Agentic recruiting

LinkedIn Hiring Asst, SeekOut, hireEZ. Application layer, not infrastructure. LinkedIn's incentive is walled garden, not open rails. These tools need ADNX more than they compete with it.

SMB HR-Tech

Factorial, BambooHR, HiBob. No agentic capabilities. No compliance vault. No network effects. Simple tools for a pre-agent world.

AI RegTech

Holistic AI, OneTrust. Horizontal, not hiring-specific. No transaction infrastructure. Compliance only — no matching, no settlement, no network.

Value creation plan

Three phases. 24 months.

M0–8

Build & Validate

Hire engineering team + compliance counsel. Launch ADNX Hire public beta in DACH. First network placements. OTP/OJP working group with ATS vendors. Founder transitions full-time post-close.

M8–16

Traction & Proof

Scale to 200 SMBs, 15+ agencies. Ship ATS integrations (Personio, Greenhouse, Recruitee). Demonstrate PLG conversion and network placement metrics. Hit €600K ARR → position for seed round.

M16–24

Scale Preparation

1,500+ SMBs. Network transaction fees compounding. Raise seed at €15M+ pre-money. Begin Nordics expansion. Build US compliance modules. Enterprise agency sales motion.

Let's talk.

If agentic infrastructure is your thesis, we'd love to connect.

adnx.ai