AI companies

Recruiting the commercial teams behind AI.

Sales, Pre-Sales, Post-Sales and Leadership across AI Infrastructure, Applications and Agents. US & Europe.

Hire GTM talentSee how we assess AI talent

Our thesis

Hiring for AI-native companies is different.

Our experience building commercial teams for AI companies has led us to a simple conclusion: the modern AI GTM hire requires a different skill set — and a different way of assessing it.

Track record still matters. But we go further: technical fluency, learning velocity, AI-native working habits, commercial translation and the ability to operate before the playbook is finished.

01 / 05The old signals

The signals that built SaaS sales teams aren’t enough anymore.

None of these four signals is worthless. Each one is useful context. The argument is narrower than that: on their own, they are insufficient predictors of success inside a modern AI company.

  1. 01

    The logo

    “They've spent five years at Salesforce.”

    A recognisable employer is useful context.

    It is not evidence of AI fluency. A strong salesperson can spend an entire career inside a mature application-software category — good training, real discipline — without ever needing the technical curiosity, learning velocity or tolerance for ambiguity that an AI-native company demands from week one.

    The question is not where somebody worked. It is what that environment required of them, and what they went looking for beyond it.

  2. 02

    The number

    “They've hit 150% three years running.”

    Historic attainment matters.

    We want the context behind it: territory, brand pull, inbound demand, installed customer base, product-market fit, SDR support, expansion versus new logo, and timing.

    A less spectacular headline number produced in an extraordinarily difficult environment can be stronger evidence than exceptional attainment inside an already-functioning commercial machine. What we are assessing is whether the performance transfers.

  3. 03

    The playbook

    “They're MEDDIC trained.”

    Process competence is useful.

    It is not a substitute for intelligence, technical-commercial fluency, adaptability, curiosity and judgement. Methodology tells you how somebody will structure a deal, not whether they can earn the conversation in the first place.

    AI buying groups include engineers, researchers, product leaders, technical founders and sophisticated executives. A certification does not tell us whether the candidate can hold a credible conversation with them.

  4. 04

    The black book

    “They already know everyone.”

    Relationships can help.

    But relationships age, buyers move, and AI keeps producing new categories and new buying groups that nobody has a network inside yet.

    We are more interested in whether somebody can create relevance than inherit it — how they get into a conversation with a technical buyer who has never heard of the company.

02 / 05Our assessment model

The assessment framework

So what do we assess instead?

Six signals of AI commercial fit.

Underneath the four surface signals sit six things we can actually investigate — built from operating and recruitment experience, not a scored test.

01

Foundational AI literacy

Can they navigate the AI stack?

A commercial hire does not need to be a machine learning engineer. They do need to orient themselves technically: where the company's product sits, what sits above and below it, and why the current wave of AI became possible at all.

That means some grasp of the role of compute, how models, infrastructure and applications interact, and why implementation considerations — data handling, latency, accuracy, integration — are the things a buyer actually worries about.

The competency is technical orientation and curiosity, not engineering expertise.

Indicative layers of the AI stack

  1. Agents

    Systems that act, not just answer

  2. Applications

    AI applied to a specific workflow

  3. Models / training

    Capability and how it is produced

  4. Compute / infrastructure

    What makes any of it possible

Layers, not a complete technical taxonomy. The competency is orientation: knowing where the product sits and what it depends on.

Layered view of the ecosystem: agents, applications, models and training, compute and AI infrastructure. Indicative layers rather than a complete technical taxonomy.

02

Applied AI fluency

Do they use AI, or merely talk about it?

Almost every candidate now claims AI fluency on the basis that they use a chatbot. On its own that is a weak signal — it is a better search engine, used well.

Have they used AI to force-multiply how they work? We are looking for people who have redesigned part of their workflow around it: research and enrichment systems, automation between tools, agents or connected workflows, relevance produced at a scale they could not reach manually.

No specific tool is a requirement. The evidence is a changed way of operating, described concretely.

AI as a tool

  1. Question
  2. Chatbot
  3. Answer

A better search engine. Useful, and a weak signal on its own.

AI as an operating layer

  1. 01Signals
  2. 02Research
  3. 03Enrichment
  4. 04Workflow
  5. 05Action
  6. 06Learning

Have they used AI to force-multiply how they work?

Comparison of AI used as a single question-and-answer tool against AI used as an operating layer running from signals through research, enrichment, workflow, action and learning.

Comparison: AI as a tool (question, chatbot, answer) against AI as an operating layer (signals, research, enrichment, workflow, action, learning).

03

Learning velocity

What have they taught themselves recently?

Experimentation, side projects, self-directed reading, building something small, breaking it, iterating, testing new tools early. Not as a personality trait — as a record.

A candidate does not need to work at an AI-native company to demonstrate this. Some of the strongest signals we see come from people inside traditional SaaS businesses who taught themselves well beyond anything their role required.

Self-directed capability over time
Rising line
Learning beyond what the role required.
Flat line
Time served in a category.

Direction, not a score. We are reading trajectory from evidence — what they built, read, tested or broke.

Trajectory of self-directed capability over time, shown as direction rather than a score.

04

Technical-commercial translation

Can they go deep without losing the business problem?

The strongest commercial people in AI hold meaningful technical detail and then translate it: into the customer's problem, the implementation consequence, the economic impact and the business outcome.

Terminology is not the goal. Nobody needs to recite hardware specifications. They need to understand enough that the technology becomes commercially meaningful to the person in front of them.

Technical depth and commercial relevance increasingly have to coexist in the same person.

Related: pre-sales and solutions engineering and solutions engineer recruitment.

From technical detail to commercial consequence

  1. 01

    Technical detail

    What the product actually does

  2. 02

    Customer problem

    Whose work it changes

  3. 03

    Implementation consequence

    What adopting it really involves

  4. 04

    Economic impact

    Why it is worth doing now

Translation chain from technical detail through customer problem and implementation to economic impact.

05

Educator mindset

Can they teach the buyer something?

The best AI sellers increasingly behave like educators. AI is still producing new categories, new workflows and new expectations, so the salesperson is often not competing for budget against a known alternative.

They are helping the customer understand what is now possible, what good implementation looks like, how frontier companies are operating, and how the customer's own organisation could work differently.

One prompt we use: teach me something about your product I don't already know. What we listen for is whether complexity survives simplification — first principles, then a clear business consequence, without the explanation becoming inaccurate.

“Teach me something about your product I don’t already know.”
Buyer range in an AI evaluation
  • ResearcherCapability and credibility
  • EngineerImplementation depth
  • ProductWorkflow and adoption
  • ExecutiveCommercial outcome
  • FounderStrategic narrative

Not every seller personally sells to all five. But AI buying groups are varied and technically sophisticated, and a seller who hands every difficult conversation to a solutions engineer loses momentum.

Complex technical concept, reduced to first principles, resolved into a clear business consequence.

06

Adaptability under ambiguity

Can they perform before the playbook is finished?

Products change mid-quarter. Positioning moves. Colleagues are more technical than the seller. Some answers do not exist yet, and the category may not have a name the buyer recognises.

We look for people who work constructively with highly technical colleagues, spar with ideas rather than defer, change their mind when the evidence changes, and stay effective without leaning on established process to tell them what to do next.

This dimension is shaped partly by running searches inside fast-moving AI companies, including our work with Unitary AI while its commercial functions were still being established.

Stages of a search environment where process is still forming.

Shareable

The hiring signal has changed.

The left column is still evidence. The right column is the additional set of questions we think AI hiring now requires.

Traditional SaaS signalRecognisable employer
What Tipped investigatesEvidence of technical curiosity
Traditional SaaS signalHistoric quota attainment
What Tipped investigatesContext behind the attainment
Traditional SaaS signalMethodology certification
What Tipped investigatesCommercial judgement
Traditional SaaS signalExisting customer network
What Tipped investigatesAbility to create relevance
Traditional SaaS signalYears in category
What Tipped investigatesLearning velocity
Traditional SaaS signalInherited playbook
What Tipped investigatesAbility to build one

Hiring for an AI company?

Discuss a search

03 / 05The interview

The Tipped interview

Questions we use to get underneath the CV.

Examples of the style of investigation rather than a fixed script. What matters is the evidence behind the answer.

  1. 01

    Teach me something about your product I don't already know.

  2. 02

    Where does your product sit in the AI stack, and what sits underneath it?

  3. 03

    Show me one way you've changed how you work because of AI.

  4. 04

    Tell me about something technical you taught yourself recently.

  5. 05

    Give me an example of a company using AI in a way that genuinely changes its economics or customer experience. Why?

  6. 06

    Tell me about a time the commercial playbook wasn't obvious. What did you do?

04 / 05AI markets

Market

Where we work in AI

AI infrastructure
Tooling, compute, platforms and the enabling layers other AI systems are built on.
AI applications
AI applied to a specific workflow, industry or customer problem, sold to the team whose work changes.
AI agents
Agentic products and systems, where the commercial story moves with the roadmap.
Frontier and emerging AI
Companies commercialising new technical capability before a mature GTM playbook exists.

The commercial profile changes across those groups, which is why we run GTM recruitment, tech sales recruitment and executive search as separate disciplines rather than one process.

05 / 05Proof

We’ve already done it.

Two searches we can describe. No numbers beyond what the clients have confirmed.

Building the commercial team behind an AI company?

Tell us what you’re hiring and where.

Hire GTM talentDiscuss a search

Common questions

Which commercial roles does Tipped recruit for AI companies?

Sales, from SDRs to founding and enterprise Account Executives; pre-sales and solutions engineering; customer success and account management; marketing; and GTM leadership up to CRO. Our searches are commercial and go-to-market rather than research or engineering.

Does someone need previous AI-company experience?

No. AI-native experience is useful evidence, but it is not the only route to it. We regularly place people from traditional SaaS businesses who demonstrated real technical curiosity, applied AI fluency and learning velocity outside what their role required.

How technical should an AI salesperson be?

Technical enough to orient themselves in the stack, hold a credible conversation with engineers and product leaders, and translate detail into a business consequence. Not an engineer. The threshold is materially higher than it was for application SaaS in the last cycle.

How does Tipped assess AI commercial candidates?

Against six dimensions: foundational AI literacy, applied AI fluency, learning velocity, technical-commercial translation, educator mindset and adaptability under ambiguity. It is a practical framework built from operating and recruitment experience, explored through evidence and worked examples rather than scored tests.

Does Tipped recruit across the US and EMEA?

Yes. We run searches in the United States and across EMEA, including first-in-region commercial hires — Cursor's first EMEA salesperson is one example.

Related expertise