Built by a recruiter who ran out of patience with the phone screen
Proofglint exists because structured verification signal is better than scheduling luck — and we built the tool that makes that true at scale.
1,200 phone screens and one calculation
Sophia Martinez spent three years as a senior recruiter at a fast-growing fintech, personally conducting over 1,200 phone screens. Midway through year two she started keeping a tally: how many of those calls revealed something a structured async flow could not have surfaced in 15 minutes?
The answer was 40%. The other 60% existed to answer questions about availability, basic role fit, and claims that were already on the resume. Sophia quit to build a tool that handled those 60% differently.
The initial prototype was a Google Form with a video-recording prompt and a manual scoring rubric. It reduced screening time by 60% in a pilot with three startup clients but collapsed under the weight of manual scoring at volume. That collapse was the insight: the tool needed to automate the scoring with AI, not just the collection.
Sophia brought in Ethan Reyes, an NLP engineer from a workforce analytics startup, and two early engineers from HireVue and Workday product teams. Proofglint now automates scoring with an AI trained on validated hiring rubrics, integrates with major ATSs, and has processed over 4,000 candidate verification flows for design-partner companies.
Make every hiring decision traceable to verified signal, not scheduling luck.
Hiring teams lose qualified candidates to scheduling friction every single day. They accept mismatched hires because interview inconsistency made meaningful comparison impossible. Proofglint addresses both of these problems by putting structured verification before the first live conversation — not as a gatekeeping step that narrows the funnel, but as a shared baseline that gives everyone on the hiring team the same starting point. When every candidate has answered the same JD-calibrated questions and every response has been scored against the same rubric, the live interview becomes a genuine conversation rather than a discovery exercise. That is a better outcome for recruiters, for hiring managers, and for the candidates themselves.
Early revenue — building with design partners
Proofglint is an early-revenue company. We are working closely with a set of design-partner talent acquisition teams at Series A through Series C companies to refine the verification scoring model, the ATS integration layer, and the candidate-facing flow. Every feature ships against a specific pain point surfaced by a real hiring team. We are not a finished product — we are a product being finished in production with people who care about getting it right.
If you are a talent acquisition manager or HR director at a growth-stage company hiring 10 or more roles per quarter and willing to give us candid feedback, we want to talk to you.
Four principles that shape every product decision
Signal over noise
Every feature we build must produce a cleaner, more actionable signal for the hiring team. If it adds volume without adding clarity, we cut it.
Fairness through structure
Structured questions and standardized scoring reduce the surface area for unconscious bias. We build toward that outcome deliberately, not by accident.
Recruiter time is precious
The goal is not to make recruiting faster in the abstract — it is to give recruiters back the hours they currently spend on calls that answer questions the async flow already answered.
Transparency in every score
Every AI confidence rating ships with the response quotes that generated it. Hiring managers see the evidence, not just the verdict.
Work with us as a design partner
We are actively expanding our design-partner program. If you are hiring 10 or more roles per quarter, we want to build Proofglint with you.