/aipeople
Embedded talent and search for AI companies US & Europe · Seed to Series B
1,000+

AI companies mapped across eleven published landscapes.

I email every founder and check to make sure I placed them correctly. That is how I know which companies just changed direction, which are about to lose people, and who is really hiring. Then I hire the commercial teams that work inside them.

The eleven parts of the AI stack I work in
A sample of the map · hover for a name · click for its landscape All eleven landscapes →
/embedded

You have a hiring plan

A dedicated hiring partner inside your tools and your process. One monthly fee, whatever the plan turns out to be, instead of a fee every time someone starts.

From $12,000 a month →
/search

You have one role that matters

A single search, run by me, on the same market and the same network. No retainer and no commitment.

20% of base, no upfront fees →
/embedded

One monthly fee, whatever the hiring plan turns out to be

You are going to be scaling the team in several areas this year. The question is how you find the right people, and what that costs.

The arithmetic

RouteYear onePer hire
In-house recruiterSalary every year, plus equity, plus severance if hiring slows215,00026,875
AgenciesA fee every time. Nothing retained between hires288,00036,000
Embedded with meOne monthly fee, whatever the hiring plan turns out to be144,00018,000

In-house recruiter, $215,000. That is a $165,000 median base, read off 16 live adverts at AI companies that published a range, plus employer taxes, benefits and a LinkedIn Recruiter licence.

Agencies and embedded are both costed at eight hires in year one on a blended base of $180,000, to compare like for like.

The reality is that embedded can fill many more roles in twelve months than eight, for the same fixed price. The more you plan to hire, the further apart those two numbers get.

The part the salary hides

A recruiter from a general tech or SaaS background arrives without a network. They do not know the twenty companies you compete with for candidates, which ones just raised, or which ones are about to lose people. They will learn. It takes three to six months, and you pay full salary throughout.

Months 1 to 3Still recruiting the recruiter. Roles sit open, or you cover them one fee at a time.
Months 3 to 6They start, and learn the market from zero. Pipeline is thin.
Months 6 to 9Strong pipelines arrive, if they are good and if they stayed.
With me, week oneLive conversations with people I already know.

Cover even two roles on a per-hire fee while they find their feet and year one is $215,000 of salary plus $72,000 of fees. That is $287,000, double the embedded price, for pipeline that starts in month six.

How it works

You are buying my time, my judgement and my network, not a process with a name on it.

  • A dedicated hiring partner inside your tools and your process.
  • Every search run by me. No juniors, nothing outsourced, no handover after the sale.
  • In your applicant tracking system, your Slack, your weekly hiring meeting.
  • A pipeline review every week, so you always see exactly where each search stands.
  • Scale up or down at each renewal point as the plan changes.
  • You keep everything: the market maps, the pipelines, the intelligence.

What it costs

Start here$8,000 One month, one role. A full market map for it and a shortlist of qualified, interested people. Credited in full against your first month if you continue.
Three months$15,000 Per month. For a burst of hiring, or to cover a gap.
Six months$13,500 Per month. The usual shape after a raise.
Twelve months$12,000 Per month. Best price and best results, because the market knowledge compounds.

US dollars, invoiced monthly in advance. No employer taxes, no benefits, no equity, no severance.

When embedded is not right

  • One or two hires and then done.You should not pay for capacity you will not use. Send me the single role that matters most and I will run it on its own, 20% of base, with no upfront fees.
  • The hiring plan is not funded yet.Come back when it is. Spend the runway on the product.
  • You already have a talent team.This still works. Extra senior capacity for a quarter, without adding permanent headcount to a plan you may need to change.
/clients

I already work in your market

Not tech in general. This market, these companies, at your stage. Twelve AI companies from Seed to Series B in the last three years.

ClearMLClearMLMLOps and AI infrastructure
CometCometMLOps and LLM evaluation
PredibasePredibaseLLM fine-tuning
TrueFoundryTrueFoundryLLMOps
HumanloopHumanloopLLM observability
ArceeArceeLLM development platform
Union AIUnion AIAI developer platform
VertesiaVertesiaAgentic process automation
QuilrQuilrAI governance
Defined.aiDefined.aiAI training data
ConfidentialMindConfidentialMindPrivate LLM deployment, sovereign AI
SimplismartSimplismartModel inference and serving

Across those companies I have placed Account Executives, Solution Engineers, Solution Architects, SDRs, Sales Managers, Product Marketing, a Director of Demand Generation, VP Sales, VP Marketing, VP Financial Services and a CRO, among others.

One of them has had seven of those searches open with me at the same time, across the US and EMEA.

/landscapes

I map this market in public

A recruiter who has worked in tech cannot show you any of this. It is the difference between knowing recruitment and knowing your industry.

1,000+
AI companies mapped, across eleven landscapes
View the maps →
~10,000
AI professionals in my network, across GTM, Product and Delivery
Connect →
11
Themes, from training data to model serving

The AI market redraws itself constantly. To stay on top of it I map a slice of it every few weeks, publish it, and then email the founders on it to ask whether I placed them correctly. They write back, at length, and often to disagree. What they say is usually more interesting than the map, and none of it is available anywhere else.

/podcast

Conversations with the people building it

Founders, CROs and CTOs from across the AI stack, on what they are building and who they need to build it.

24 episodes · scroll the list Listen to the episodes →
  • NeuralkNeuralkTabular & Structured-Data MLAlexandre PasquiouCo-Founder & CSO
  • V7V7Document Processing & ExtractionAlberto RizzoliCo-Founder & CEO
  • Reality DefenderReality DefenderDeepfake & Content Authenticity DetectionBen ColmanCo-Founder & CEO
  • WitnessAIWitnessAIShadow AI Discovery & Workforce AI GovernanceRick CacciaCEO
  • AsenionAsenionAI Governance, Risk & ComplianceAnna FelländerCEO
  • Adversa AIAdversa AIAI Red Teaming & Security TestingAlex PolyakovCo-Founder & CEO
  • HammingHammingLLM Evaluation & TestingSumanyu SharmaCo-Founder & CEO
  • Align AIAlign AIAI Governance, Risk & ComplianceNicholas WooHead of CX & Growth
  • TraceloopTraceloopLLM & Agent ObservabilityNir GazitCEO & Co-Founder
  • InsightFinderInsightFinderLLM & Agent ObservabilityHelen GuFounder & CEO
  • NebulyNebulyAgent Monitoring & DebuggingJulien RouxCo-Founder
  • ConfidentialMindConfidentialMindAI Data Privacy & Confidential ComputingMarkku RäsänenCEO
  • Contextual AIContextual AIEnterprise Knowledge & Context LayerDouwe KielaCEO
  • Pruna AIPruna AIInference Optimization & CompilersQuentin SinigGTM Leader
  • OpenPipeOpenPipeModel Training & Fine-TuningKyle CorbittFounder
  • AnoteAnoteData Labelling & Annotation ServicesNatan VidraFounder & CEO
  • VectorizeVectorizeAgent Memory & Context ManagementChris LatimerCo-Founder & CEO
  • Feedback IntelligenceFeedback IntelligenceAgent Monitoring & DebuggingChinar MovsisyanFounder
  • Arcee AIArcee AIFoundation Model LabsBrian BenedictCo-Founder & CRO
  • LangWatchLangWatchLLM Evaluation & TestingManouk DraismaCo-Founder
  • DynamiqDynamiqAgent & Workflow Orchestration FrameworksVitalii DukCEO & Founder
  • KorticalKorticalAutoML & Predictive AnalyticsBarbara JohnsonCOO
  • PromptLayerPromptLayerPrompt Management & ExperimentationJared ZoneraichFounder
  • ValohaiValohaiMLOps & ML Lifecycle PlatformsEero LaaksonenCEO
/posts

What I publish

Market maps, hiring notes, and what founders tell me when they disagree with me.

Most of it goes out on LinkedIn. The maps are the biggest pieces, built on what the companies on them wrote back when I asked whether I had placed them correctly. Several disagree with me in public, which is the most useful thing that can happen.

Alongside those, shorter notes on who is hiring, which roles are moving, and what I am seeing across the AI stack week to week.

Follow along on LinkedIn →

/start

The next step