ApexMake vs Ethos
Both companies point voice AI at expert interviews. What happens next is where they split: a better introduction, or an answer.
- The AI interview: builds a richer profile, against a queryable knowledge base
- Delivery: a matched expert, against a cited answer
- Focus: a broad marketplace, against investment diligence
The short answer
- Ethos is an AI-native expert marketplace founded in 2024 by James Lo, formerly of McKinsey and SoftBank, and Daniel Mankowitz, formerly of Google DeepMind. It raised a $22.75M Series A led by a16z in May 2026, taking total funding to roughly $30M.
- Its mechanic is voice onboarding. An AI agent runs an extended interview with each expert to capture sub-specialisations a job title would miss, supplemented by ingesting their published work. Around 35,000 experts join weekly, mostly by invitation.
- Ethos monetises the introduction, with a per-project take rate reported at 30% or more, and serves hedge funds, private equity firms, AI labs and consulting firms across a wide range of expertise.
- ApexMake starts from the same insight and goes one step further. The AI interview is not there to build a better profile, it is there to build a queryable knowledge base, so the client gets a cited answer in roughly six seconds instead of a shortlist and a call to schedule. The call still exists, it is just optional, and booking one is instant because the expert was vetted on the way in.
- Ethos is the better fit when you need to find and hire the right specialist across many domains. ApexMake is the better fit when you need the answer itself, in a form you can cite, on a deal deadline.
Which should you use?
Ethos and ApexMake are the two clearest expressions of the same thesis, that AI should rebuild expert networks from the inside rather than decorate them, and they apply it to different halves of the problem. Ethos applies AI to discovery: voice interviews produce far richer expert profiles than a CV, and matching against them beats keyword search over a database by a distance. ApexMake applies AI to delivery: the interview captures the substance of what the expert knows, so the client queries it directly. Bottleneck is finding the right person across any domain? Ethos. Bottleneck is that finding them still leaves you a call away from the answer? ApexMake.
What sets ApexMake apart
| Dimension | ||
|---|---|---|
| What the AI interview produces | A queryable knowledge base of what the expert knows | A richer expert profile for matching |
| What the client receives | A cited answer with follow-ups, plus a direct line to book the expert live | A matched expert, then a call or project |
| Time to insight | About 6 seconds (median) | Much faster matching than a traditional network, still gated by the call |
| Coverage strategy | Deep by design: AI infrastructure, inference, training, 1,000+ software and infra products | Broad by design: expertise across many domains, including hiring and consulting use cases |
| Network growth | Selective recruitment, screened for senior first-hand operating experience | Around 35,000 experts joining weekly, mostly invitation-based |
| Pricing shape | One subscription covering corpus answers and expert calls | Pay-as-you-go per project, reported take rate of 30% or more |
| Primary buyer job | Answer an investment question before the meeting | Find and engage the right specialist for a piece of work |
| Compliance posture | MNPI screening and a timestamped audit trail on every query, built for regulated capital | Marketplace model serving funds, labs and consultancies across mixed use cases |
| Expert economics | Experts earn when accepted knowledge helps clients, not only when they are free for a call | Experts are paid per engagement, with the platform taking a project fee |
What Ethos is
- 2024, by James Lo (ex-McKinsey, ex-SoftBank Vision Fund) and Daniel Mankowitz (ex-Google DeepMind)
- AI-native expert marketplace. Voice agent onboarding and AI matching, monetised per project.
- $22.75M Series A led by a16z in May 2026, with General Catalyst, XTX and others. Roughly $30M total across two rounds.
- Around 35,000 experts joining weekly, largely by invitation. A small team reported at about eight people, tracking toward eight-figure annualised revenue.
- Pay-as-you-go per project, with a reported take rate of 30% or more.
- Hedge funds, private equity firms, AI research labs and global consulting firms.
Same insight, different half of the problem
Ethos spotted something correct and important: a job title is a terrible description of what somebody actually knows. Two people with identical LinkedIn headlines can have completely different operating experience, and any matching system built on titles will keep making the same mistakes. So Ethos points a voice agent at the expert and lets them talk, capturing sub-specialisations no structured field would have held, then enriches that with their published work.
ApexMake starts from the same observation and asks the obvious follow-up. If the AI is already running a deep interview, why should the output stop at a better profile?
It should not. Once you are running a structured, thorough interview you can capture the substance too: how the thing actually works, what the real numbers were, where the expert's view parts company with consensus. Then you make that directly queryable. The profile improvement is a side effect. The knowledge base is the product.
Matching is not the bottleneck for a deal team
For a recruiter or a consulting firm staffing a project, discovery really is the bottleneck. Finding somebody who worked on finance automation at a funded startup is hard, and a system that answers that query in natural language is a big improvement.
For a deal team three days from an investment committee, discovery is not the bottleneck. Delay is. Being handed a perfectly matched expert on Tuesday afternoon does nothing if the call lands Thursday and the memo is due Wednesday night.
Which is why the two products feel different in use despite sharing a technical foundation. Ethos makes the introduction dramatically better. ApexMake removes the need for one in the common case, because the interview already happened and the knowledge is sitting there. And when a deal team does want the person live, the booking is immediate: the expert was vetted on the way in, and their previous answers show you exactly who you are getting.
Breadth of a marketplace versus depth of a network
Ethos is scaling fast and wide. Roughly 35,000 experts join each week, largely by invitation, spanning hedge fund research, AI labs, consulting and hiring use cases. That breadth is a strategy, and it makes the marketplace useful to a lot of different buyers.
ApexMake takes the opposite bet. Rather than maximising network size, it screens hard for senior first-hand operating experience and concentrates on AI infrastructure, inference, training and the software and infrastructure products investors keep underwriting. In those domains the constraint is never the number of profiles. It is whether the person answering has run the system in production.
Depth compounds, too. Because every ApexMake expert is interviewed in depth rather than lightly profiled, each addition to the network is worth more for the specific questions the network exists to answer.
Take rates versus subscriptions
Ethos monetises transactions: pay-as-you-go per project, with a reported take rate of 30% or more. That suits irregular, project-shaped demand well. You pay when you engage somebody and nothing when you do not.
It carries the property every transactional model carries. The marginal question has a price on it, and teams with high, unpredictable question volume end up budgeting their curiosity.
A subscription to a knowledge base makes the marginal question effectively free, which is the behaviour change that matters most in diligence. The tenth follow-up, the one that surfaces the problem, costs nothing to ask.
What regulated capital needs on top
Serving hedge funds and private equity is not the same job as serving consultancies and AI labs. Regulated capital needs material non-public information screened on every interaction, and needs to prove later that it was.
ApexMake is built around that constraint rather than adapted to it. Every query is screened before it runs, every expert match carries a conflict check, every answer is logged with its citations, every follow-up is screened again, each event timestamped into an audit trail.
That is a narrower product decision than a general-purpose expert marketplace can make, and it is intentional. Compliance is the foundation rather than a feature, because for a fund an answer that cannot be defended is worse than no answer at all.
Choosing between them
Ethos fits better when
- Your needs span many domains at once, from expert hiring to consulting staffing to research, rather than concentrating in investment diligence.
- Discovery really is your bottleneck, and you need people whose relevant skills a job title would never reveal.
- Project-shaped, irregular demand suits pay-as-you-go pricing better than a subscription.
- You want the widest possible pool of invited experts and are comfortable that the engagement ends in a scheduled conversation.
ApexMake fits better when
- You need the answer, not the introduction, and you need it inside the meeting.
- Your questions concentrate in AI infrastructure, inference, training and the surrounding software and infra stack.
- The output has to be citable in an IC memo, with each claim traceable to a named expert.
- Compliance requires MNPI screening and a machine-readable audit trail on every interaction.
- Your team's question volume makes per-project pricing a limit on how much you actually ask.
- You want the call to be optional rather than the product, and instant to book when you do want it.
ApexMake vs Ethos, explained
What does Ethos do?
Ethos is an AI-native expert network. A voice agent conducts an extended onboarding interview with each expert to capture skills and sub-specialisations beyond their job title, supplemented by ingesting their published work. Clients describe what they need in natural language, get matched to experts, and pay per project with a reported take rate of 30% or more.
How is ApexMake different if both use AI interviews?
The difference is what the interview is for. Ethos uses it to build a richer profile so matching improves. ApexMake uses it to capture the substance of what the expert knows, so clients query that knowledge directly and get a cited answer in about six seconds rather than a shortlist and a scheduled call. Calls remain available, booked directly with the same pre-vetted experts.
Is Ethos an investor-focused product?
Partly. Its clients include hedge funds and private equity firms, but also AI research labs, consulting firms and hiring use cases. ApexMake is built specifically for investment research and diligence, which is why MNPI screening and audit logging are core rather than optional.
Which network is larger?
Ethos is scaling much faster in headcount terms, with roughly 35,000 experts joining weekly. ApexMake optimises for depth over size, screening for senior first-hand operating experience and concentrating coverage in AI infrastructure and the software and infra stack.
How much does Ethos cost?
Ethos runs a pay-as-you-go transactional model with a per-project fee reported at 30% or more of the engagement. Specific rates are negotiated per project rather than published.
Sources
Figures describing Ethos are drawn from the company’s own published materials and third-party reporting, and were accurate as of the date of publication. Product and pricing details change; verify with the provider before making a purchasing decision.