Clera

Clera is a two-sided recruiting service built around an AI “talent agent” rather than a conventional job board. On the candidate side, the user describes the role they want, preferences and dealbreakers; Clera then searches its company network, surfaces relevant opportunities and can make direct introductions instead of sending the candidate through a generic application funnel.¹ The product’s strongest idea is not automatic résumé writing or keyword matching. It is reducing the search-and-apply workload by turning candidate intent into a continuously running recruiting process.

The employer proposition mirrors that design. Clera markets access to candidates whose intent and preferences have already been collected, with profiles matched to a role’s required skills, stack and culture and introductions delivered directly rather than through a large inbound applicant queue.² The employer page says relevant candidate profiles can be delivered within roughly 48 hours and emphasizes direct introductions, including Slack-based workflows.² These are vendor service claims rather than independently measured time-to-fill guarantees, but they describe a materially different workflow from an ATS or sourcing database.

Clera is explicit that AI should assist rather than decide. Its AI disclosure page frames the system as “Humans Decide, AI Assists”: AI handles matching, research, interview preparation, CV/profile analysis, job understanding and candidate-company coordination, while candidates and companies make the actual employment decisions.³ The same page says the system is model-agnostic and may route work across Claude, GPT and Gemini models. Clera is therefore best assessed as a recruiting application and orchestration layer, not as a proxy for the quality of any upstream foundation model.

The matching layer creates structured candidate information. Under the current Terms, Clera may generate a candidate profile from information such as a résumé, LinkedIn data, calls and other interactions, and may produce assessments, summaries or match scores for use in the recruiting workflow.⁴ Potential employers can receive the candidate’s professional information and AI-generated profile when Clera makes a match or introduction.⁴⁵ This can reduce repeated screening work, but it also means the quality and governance of the generated profile matters because the artifact is not merely private coaching text: it can shape how an employer sees the candidate.

The AI stack is unusually transparent for a recruiting product. Clera publicly identifies tools including the Vercel AI SDK for model routing, Trigger.dev for background workflows, Langfuse for tracing, Typesense for retrieval, Zod for structured outputs, automated evaluations and manual review.³ It says successful outcomes improve the system through better prompts, context and evaluation rather than through retraining the underlying foundation models.³ That architecture disclosure does not independently prove matching quality or fairness, but it makes the system easier to reason about than a generic claim that “AI finds the right candidate.”

The candidate AI call is optional under the Terms.⁴ That is important because the service can collect richer conversational data when a candidate chooses to interact by call, but a candidate is not contractually required to accept an AI call simply to use every part of the service. Clera’s policy still permits authorized staff to review conversations, email, LinkedIn and behavioral data for service operation and matching, so the system should be understood as an assisted recruiting service with human-accessible operational data rather than a fully private self-service agent.⁴

Clera also has a recruiting-partner channel. Independent recruiters can submit candidates into the network without paying membership/software fees and, when a candidate is hired, Clera says the recruiter can receive 10% of the candidate’s first-year base salary.⁶ That expands sourcing beyond candidates who discover Clera directly, but the public scale figures on the partner page are not perfectly synchronized across current copy. The inconsistent figures are not reliable enough to support one exact current employer-count claim.

The main product boundary is that Clera is not a full ATS or universal job marketplace. Its value depends on the quality and breadth of its employer network, its ability to interpret candidate intent and the trust candidates place in its outreach. The product is most differentiated when it can replace dozens of low-probability applications with a smaller number of relevant warm introductions; it is much less differentiated if the match layer simply produces another automated message about a generic role.

Candidates currently use Clera for free. The public pricing page says there is no candidate charge and no card requirement.⁷ This creates unusually clean economic alignment on the candidate side: the person looking for work does not need to decide whether another subscription is worth paying while unemployed or exploring the market.

Companies pay on successful placement. Clera’s standard employer fee is 15% of the hired candidate’s first-year base salary, with no upfront or monthly software fee.⁷⁴ The Terms describe payment in installments at 30, 60 and 90 days after the start date and a 90-day replacement or prorated-credit guarantee if the placement does not work out.⁴ That makes Clera economically closer to a contingency recruiting agency than to an ATS, sourcing SaaS subscription or job-board posting fee.

For contract or fractional placements, the pricing page says Clera charges 15% over the first 12 months rather than a one-time permanent-placement calculation.⁷ Employers therefore need to model cost differently depending on employment structure.

Clera compares its 15% success fee with traditional agency fees of roughly 20–30% plus, in some cases, retainers.⁷ That comparison is vendor positioning rather than an independent market benchmark; the current 15% fee is the reliable pricing fact, while the competitor range is context only.

The contingency model has a clear advantage: an employer can evaluate the service without paying before a hire. It also has a clear cost characteristic. On a $150,000 base salary, for example, the standard 15% fee would be $22,500; on a $200,000 base salary it would be $30,000. The percentage can therefore be economical compared with a traditional agency but materially more expensive than self-serve recruiting software if an internal team could have sourced the hire efficiently itself.

The candidate/employer incentives are also asymmetric in a useful way. Candidates pay nothing, while Clera earns revenue only when a company hires someone. That rewards completed placements rather than candidate engagement time. It does not, however, automatically guarantee that every outreach message is well targeted: a success-fee business can still have incentives to broaden top-of-funnel prospecting.

Recruiting partners have a separate revenue-sharing structure: the public partner page says there is no membership/software fee and that a partner recruiter can receive a 10% first-year-base payout when a submitted candidate is hired.⁶ This is not part of the ordinary employer fee, but it helps explain why Clera can source candidates from both direct users and external recruiters.

Overall, pricing is now one of the clearer parts of the product. The older concern that commercial limits were opaque is no longer accurate. The more important buyer questions are whether the network produces enough relevant hires to justify a 15% contingency fee and whether the outreach/matching experience is trusted by candidates.

Clera has a more meaningful current user-feedback corpus than many young AI recruiting tools. Trustpilot currently shows 3.9/5 from 69 reviews, with 77% five-star and 20% one-star reviews.⁸ The distribution is highly polarized rather than simply mediocre. Positive reviewers frequently describe relevant matches, quick interview introductions, easy onboarding and helpful support; some report successful job outcomes. Negative reviewers repeatedly focus on unsolicited or persistent outreach, uncertainty over whether the recruiter communicating with them is human or automated, use of publicly available professional information before signup, missing context about roles and concerns over consent or data handling.⁸

The 3.9/5 Trustpilot score should not be read as a direct measure of Clera’s AI quality. Trustpilot is a self-selected review surface, and the reviews mix the quality of matches, outreach, customer support and the broader recruiting service rather than isolating one AI capability. The sample is nevertheless large enough to be a serious qualitative signal: Clera’s main external risk is not a lack of people saying the product works, but a split between users who value proactive matching and recipients who dislike how that proactivity is delivered.

Product Hunt is much weaker as rating evidence. Clera launched there in April 2026 and currently shows 5.0/5 from only one review.⁹ The launch reached the top product position for the day and generated substantial launch engagement, but one rating is not statistically meaningful. It is useful for launch and traction context only.

Clera’s own April 2026 funding announcement reported a $3 million pre-seed round and, at that time, more than 80,000 users and 600 startups in the network.¹⁰ Those are historical first-party figures, useful for showing that the product had real distribution around launch rather than being a landing-page prototype. They are not treated as current independently verified scale metrics.

No current exact-product G2, Capterra, TrustRadius or comparable structured enterprise-review corpus was found. The external evidence base is therefore meaningful but concentrated, with Trustpilot providing the principal current exact-product user-feedback surface alongside first-party documentation.

The most important trust issue is Clera’s pre-onboarding outreach model. The Privacy Policy explicitly says Clera may contact professionals who have never created an account, using information such as name, job title, employer and other publicly available professional details, with legitimate interest as the stated legal basis for prospecting.⁵ For those people, the Privacy Policy governs the processing before they accept the Terms.⁵ This disclosure is directly relevant to the negative Trustpilot pattern: unsolicited outreach is not merely anecdotal behavior; it is part of the documented acquisition model.

That does not establish that the practice is unlawful or improper in a particular jurisdiction. It does mean users should understand that Clera is not only a tool people voluntarily open after signup. It can proactively identify and contact potential candidates based on public professional data, and recipients who dislike that model may experience the service very differently from candidates who intentionally join to receive opportunities.

Once someone uses the service, Clera collects a broad recruiting-data set: identity/contact information, résumé and LinkedIn information, work and education history, skills, interaction data, AI conversations, calls, recordings or transcripts, and AI-generated profiles, summaries and match scores.⁵ The service can use those data to provide matching and introductions, operate and improve its AI systems, communicate with users, market the service, prevent abuse and meet legal obligations.⁵

Clera lists several important third-party service categories and providers, including Anthropic, OpenAI and Google for AI models; Vapi and ElevenLabs for voice-related functions; Microsoft, Supabase, Vercel and Brevo for other infrastructure or communication functions.⁵⁴ The current policy says an up-to-date subprocessor list can be requested rather than publishing the complete live list directly. For enterprise or regulated buyers, an on-request list is workable but less convenient than a continuously public subprocessor register.

Data retention is comparatively concrete. Clera says it generally keeps personal information for two years after the last recorded interaction and, when a valid deletion request is made, removes information from active systems within 30 days, subject to stated legal/operational exceptions.⁵ The policy also describes international transfers, including to the United States, and use of safeguards such as Standard Contractual Clauses where applicable.⁵

Clera uses PostHog session replay with data masking according to the Privacy Policy.⁵ “Masked” is important; it is not equivalent to “no behavioral recording.” Candidates who are especially sensitive to recruiting-data capture should evaluate the full policy rather than assuming the candidate-facing free service means minimal data collection.

Automated-decision governance is another meaningful area. Clera says AI does not make the final hiring decision and provides a path to request human review when an automated decision materially and negatively affects an opportunity.³⁵ The Terms also address New York City’s automated-employment-decision-tool framework, state that Clera conducts bias audits where applicable and say summaries are made publicly available.⁴ A Clera-specific public bias-audit summary was not located in the available current evidence. That is a documentation-discoverability gap, not evidence that no audit exists. A buyer relying on the representation should request or locate the current audit summary before deployment in a regulated hiring process.

Clera also operates a public bug-bounty/responsible-disclosure program and says it responds to valid high-risk vulnerability reports within five business days.¹¹ This is a positive first-party security-transparency signal. It is not an independent penetration test, SOC report or security certification.

The available evidence supports moderate confidence rather than high confidence. Clera has enough current exact-product user feedback to identify both real success cases and recurrent trust failures, and its legal/AI documentation is unusually specific. It still lacks a broad independent enterprise-review corpus, controlled matching benchmark, independently verified fairness evidence or external assurance sufficient to convert the product into a high-confidence quantitative rating.

Clera is best suited to candidates who want startup roles and prefer curated introductions to the repetitive job-board process. A candidate who can clearly state role, seniority, location, compensation, stack, company-stage and other dealbreakers benefits more from the product than someone who simply wants to browse every possible opening.¹ The service is designed to do search and matching work continuously rather than wait for the candidate to submit dozens of applications.

It is particularly attractive when the candidate values warm introductions. Clera’s employer workflow is built around direct introductions rather than throwing another application into an ATS queue.² Current positive Trustpilot reviews suggest that this can translate into fast conversations and, in some cases, successful placements.⁸ Those reports are not a guarantee of match quality for every profile, but they support the core use case.

On the employer side, Clera fits growth-stage startups and other companies hiring technical or GTM talent that are comfortable with a contingency recruiting fee. The absence of upfront/monthly fees lowers adoption friction, while the 15% first-year-base charge aligns payment with an actual hire.⁷ Companies with very strong internal sourcing teams should compare that fee with the marginal value of Clera’s network and screening rather than treating the zero-upfront structure as zero cost.

Clera is less suitable for candidates who strongly object to proactive recruiting outreach or to public professional information being used before they create an account. That is a documented part of the prospecting model, and the external review corpus shows that it is one of the service’s most contentious characteristics.⁵⁸

It is also a weaker fit for users uncomfortable with AI-assisted profiling. Candidate data can be summarized into AI-generated profiles and match assessments that are shared with potential employers as part of the recruiting workflow.⁴⁵ A user who wants an entirely manual recruiter relationship or who does not want third-party model providers processing relevant interaction data should not assume Clera operates that way.

Regulated employers should perform additional diligence before using Clera as part of an employment-decision process. The Terms contain explicit AEDT/bias-audit language, but a Clera-specific public audit summary was not located in the available current evidence.⁴ Companies subject to NYC Local Law 144 or similar rules should obtain the actual applicable audit/notice materials and understand which part of their workflow Clera performs.

Finally, Clera is not the obvious choice for employers primarily seeking an ATS, interview-scheduling system or broad HRIS suite. Its value is sourcing, matching and introduction. A company that already has strong candidate flow but needs downstream workflow management may need a different product category.