Ethos has raised $22.75 million in a round led by a16z to scale what it describes as an expert network platform built around voice onboarding. The company’s pitch is straightforward: if you want to match decision-makers with credible specialists—whether for research, product strategy, technical evaluation, or other high-stakes work—you can’t treat onboarding like an afterthought. You need a system that gets the right people into the network quickly, consistently, and with enough signal to make the resulting recommendations useful.
What makes this round notable isn’t only the size of the check—though $22.75 million is meaningful for a company operating in a category that often requires both operational rigor and trust-building. It’s the emphasis on voice as the primary onboarding mechanism, paired with a striking throughput claim: Ethos says it is onboarding 35,000 experts per week.
That number alone suggests Ethos is not trying to build an “invite-only” directory where onboarding is slow and curated by hand. Instead, it appears to be pursuing a more industrial approach to expert acquisition—one that treats onboarding as a repeatable workflow rather than a bespoke process. Voice, in this context, is less about novelty and more about speed and structure: a way to capture expertise signals from people who may not have the time or inclination to fill out long forms, upload documents, or navigate complex qualification steps.
The broader market context matters here. Expert networks have existed for years, but the demand pattern has shifted. Enterprises increasingly want rapid access to domain knowledge, often on tight timelines and with evolving questions. In parallel, AI-driven workflows are changing how organizations consume expertise: instead of waiting for a human to manually interpret a problem, teams are increasingly using tools that can summarize, compare, and evaluate information—then asking humans to validate, refine, or challenge outputs. That creates a new kind of “expert availability” problem. It’s not just about having experts; it’s about having experts who can be engaged quickly and reliably, and whose credentials can be verified at scale.
Ethos’ bet is that voice onboarding can reduce friction while improving the quality of the onboarding data. In practice, voice can serve as a natural interface for experts to describe their background, experience, and areas of competence. It also allows the system to capture nuance—how someone explains tradeoffs, how they talk through methodology, and how they respond when asked to clarify. Compared to text-only intake, voice can feel more conversational, which may increase completion rates and reduce drop-off. Compared to purely manual screening, voice-based workflows can be standardized and accelerated.
The company’s framing—“voice onboarding as a core product approach”—signals that Ethos is treating this as a foundational capability rather than a feature. That matters because onboarding is where many expert networks either win or lose. If onboarding is too slow, the network becomes stale. If onboarding is too shallow, the network becomes noisy. If onboarding is too expensive, the business model breaks under growth. Ethos is effectively claiming it can thread that needle by using voice to gather structured signals at high volume.
A unique angle in Ethos’ approach is the implied shift from static credentialing to dynamic qualification. Traditional expert directories often rely on resumes, self-reported profiles, and periodic updates. Those can work, but they tend to be limited in how quickly they reflect current expertise and how well they map to specific use cases. Voice onboarding, by contrast, can be designed to elicit information that is more directly relevant to the kinds of engagements customers want—such as evaluating technical approaches, assessing market dynamics, or advising on product decisions.
Even without seeing the internal mechanics, the throughput claim suggests Ethos has built a pipeline that can handle large volumes without sacrificing consistency. Onboarding 35,000 experts per week implies automation across multiple stages: intake, identity verification, qualification prompts, categorization, and likely some form of quality control. Voice can be integrated into that pipeline in a way that supports both human review and machine-assisted processing. For example, the system can transcribe and analyze responses, then route experts into appropriate categories or flag uncertain cases for additional checks.
This is where the a16z-led investment becomes more than a funding event—it’s a validation of a thesis about how expert networks should evolve. a16z has been vocal about the importance of infrastructure for AI-era workflows, and expert networks are increasingly positioned as a form of “human-in-the-loop” infrastructure. But infrastructure only works if it scales. Ethos’ emphasis on onboarding speed suggests it is building the scaling layer that makes expert matching viable at enterprise pace.
There’s also a subtle but important implication in Ethos’ focus on voice: it acknowledges that expertise is often communicated best in conversation. Many experts can write a bio, but fewer can translate their experience into the specific language that customers need when they’re trying to make a decision. Voice onboarding can prompt experts to explain their work in a way that reveals depth and relevance. It can also help the system detect whether someone’s claims align with how they describe their experience—an important consideration for trust.
Trust is the currency of expert networks. Customers pay for credibility, not just availability. If Ethos is onboarding tens of thousands of experts weekly, it must have a way to maintain quality. That doesn’t necessarily mean every expert is fully vetted by a human reviewer before being added. It could mean Ethos uses a layered approach: automated screening for baseline qualification, targeted human review for edge cases, and ongoing monitoring based on engagement outcomes. The key is that quality assurance must scale alongside onboarding volume.
Ethos’ positioning also reflects a broader shift in how companies think about “expertise.” In many organizations, expertise used to be centralized—stored in internal SMEs, consultants, or long-standing vendor relationships. Now, expertise is increasingly distributed and time-sensitive. A customer might need a specialist for a narrow question, a specific technology evaluation, or a market assessment that changes quickly. That means the network must be both broad enough to cover many domains and precise enough to deliver relevant matches.
Voice onboarding can support both goals. Broad coverage comes from reducing friction for experts to join. Precision comes from capturing richer signals during onboarding—signals that can be mapped to customer needs. If Ethos can translate voice responses into structured attributes—such as domain categories, depth indicators, methodologies, and confidence levels—it can improve matching accuracy. Better matching reduces wasted engagements, which in turn improves customer satisfaction and retention.
The company’s reported onboarding rate also hints at a potential competitive advantage: speed. In expert networks, speed isn’t just about getting experts into the system; it’s about getting customers answers faster. When onboarding is slow, customers experience delays that can undermine trust. When onboarding is fast, the network can respond to urgent requests and emerging needs. That can be especially valuable in environments where decisions are time-bound—such as product launches, regulatory assessments, security evaluations, or rapid research cycles.
At the same time, speed introduces risk. Rapid onboarding can lead to lower signal quality if the process is too permissive. Ethos’ voice-first approach suggests it is trying to mitigate that risk by making onboarding more informative than a simple form submission. A voice interaction can be designed to test understanding, not just collect claims. It can also reveal communication clarity, which can correlate with how useful an expert will be during real engagements.
Another interesting dimension is how voice onboarding might change the economics of expert networks. Many expert networks struggle with the cost of onboarding and the cost of maintaining the network. If onboarding requires extensive manual work, scaling becomes expensive. If onboarding is too automated without quality controls, the network becomes unreliable. Ethos’ approach implies a model where automation handles the bulk of onboarding, while quality assurance is applied selectively. That can lower marginal costs and make growth sustainable.
The investment also raises questions about what Ethos will do next. Scaling onboarding is one part of the equation; scaling engagement quality is another. As the network grows, Ethos will need to ensure that experts are not only qualified but also responsive and aligned with customer expectations. That includes scheduling, communication norms, and the ability to handle different engagement types—some of which may require deeper collaboration than a short Q&A.
Voice onboarding could play a role beyond initial qualification. It could become a standard interface for ongoing interactions, such as follow-up clarifications or additional screening when customers request experts for specialized tasks. If Ethos builds a consistent voice-based experience, it can reduce friction for both experts and customers, potentially increasing conversion rates from onboarding to active participation.
There’s also the question of how Ethos differentiates itself from other expert network models. Some networks focus on curated talent and long-term relationships. Others emphasize breadth and rapid matching. Ethos appears to be aiming for a hybrid: broad onboarding at scale, paired with a mechanism designed to preserve quality through structured voice intake. The unique take here is that Ethos is not simply adding voice as a user interface improvement; it’s using voice as a way to generate better onboarding data and accelerate the pipeline.
In a world where AI systems can already transcribe and analyze speech, voice onboarding becomes a bridge between human expertise and machine-readable structure. That bridge is crucial for expert networks because matching and routing depend on structured attributes. If Ethos can convert voice responses into reliable signals—whether through transcription plus analysis, rubric-based scoring, or other methods—it can improve the network’s ability to find the right expert for the right question.
The company’s claim of onboarding 35,000 experts per week also suggests it has built operational capacity to manage that volume. That likely includes tooling for identity verification, fraud prevention, and compliance considerations. Expert networks can be vulnerable to misrepresentation, especially when onboarding is scaled. Voice onboarding can help detect inconsistencies, but it doesn’t eliminate the need for robust verification. Ethos’ ability to scale implies it has invested in these operational layers.
From a customer perspective, the value proposition is likely centered on reduced time-to-expert and improved match quality. Enterprises don’t just want “an expert”; they want an expert who can answer the question they actually have, in the