SageScreen vs. Harver

The Platform You Build vs. The Platform You Buy

If you just read our comparison with Pymetrics, you already know half this story. Pymetrics — the neuroscience-based hiring tool built by a Harvard/MIT scientist — was acquired by Harver in August 2022. To understand what that meant for Pymetrics, we had to understand what Harver actually is.

Turns out, that’s a more interesting question than it sounds.

Harver isn’t a product that grew into a platform. It’s a private equity acquisition strategy that assembled a platform out of purchased parts. The company that calls itself Harver today started as something called Outmatch, backed by Rubicon Technology Partners and Camden Partners. It completed seven acquisitions in four years, took the name of one of the companies it bought, and now markets the assembled pieces as a “comprehensive volume hiring solution.” It has had three CEOs since 2021.

None of that means Harver’s tools don’t work. They have 1,300+ customers, including McDonald’s, Booking.com, and Valvoline. They’ve processed 100 million candidates. The I/O psychology underpinning their assessments is rooted in 35 years of data. That’s real.

But there’s a difference between a platform that was built with a coherent vision and one that was assembled through acquisitions to create a comprehensive-looking product line. That difference shows up in the product, the experience, and — if Glassdoor reviews from their own employees are any indication — the trajectory.

The Acquisition Timeline

Understanding what Harver is requires understanding how it was assembled. This isn’t ancient history — the entire construction happened in under four years.

That’s seven acquisitions, three CEOs, one rebrand, and a name borrowed from an acquired company — all in four years. The original Harver was an Amsterdam startup founded by Barend Raaff. The company called “Harver” today is a Dallas-based PE portfolio company wearing an Amsterdam brand.

What Harver Actually Offers

Harver markets itself as an end-to-end volume hiring solution. To understand the product, it helps to understand which acquisition each piece came from.

Each of these tools has genuine capability. I/O psychology-validated assessments are real. The Pymetrics neuroscience games are peer-reviewed. Checkster actually does automate reference checking. The question isn’t whether individual pieces work. The question is whether an assembled platform behaves like a coherent product — and whether the roadmap serves the tools or the portfolio strategy.

Where SageScreen Fits: A Different Category

Harver is a pre-screening assessment platform. SageScreen is an AI interviewing platform. These are different things, and it matters to be precise about the distinction.

Harver’s tools sit before the interview in the hiring funnel. They filter, sort, and rank. After Harver does its work, someone still has to interview the candidates who passed. SageScreen is the interview — the part where a candidate’s competency, communication, and judgment are actually evaluated.

The Video Interview Problem

Harver does include a “virtual interview” module, which is where the category overlap gets confusing. But Harver’s own documentation describes it clearly: it’s a one-way, asynchronous video recording tool. Candidates record answers to preset questions. Recruiters watch the recordings later. There is no AI evaluation, no adaptive follow-up, no automated scoring of responses.

Harver’s knowledge base is explicit: their video interview module “does not make use of automated decision-making or facial and/or micro-expression analysis technology.” It’s a qualitative supplement to their quantitative assessment data — a way for recruiters to get a sense of candidates before committing to a live interview. It is not, by Harver’s own definition, meant to replace the face-to-face interview.

This isn’t a criticism of Harver’s video tool for what it’s designed to do — replacing phone screens with asynchronous video is a reasonable efficiency gain. But it’s not the same category as an AI that conducts and evaluates an actual interview. The gap between “record a video for someone to watch later” and “have an adaptive conversation that produces a structured evaluation” is the gap between a voicemail and a phone call.

What Harver Produces vs. What SageScreen Produces

The most practical comparison is output. What does the hiring team actually receive?

The Built vs. Bought Architecture

Assembled platforms face a structural challenge: each acquired module was built for a different purpose, by a different team, with a different technical architecture. Integration is possible but never seamless. The seams show up in candidate experience (different interfaces for different assessment modules), in data flow (results that don’t naturally connect across tools), and in roadmap priority (which acquired module gets engineering attention this quarter?).

SageScreen was built as a single system from the start. The 10 AI agents across 3 isolated pipelines — conversation, evaluation, and reporting — were designed to work together. The conversation engine doesn’t know the score. The scoring engine doesn’t influence the questions. The report generator draws from both but can’t modify either. These aren’t acquired tools stitched together; they’re a single architecture with deliberate separation of concerns.

The Pricing Comparison

Harver does not publish pricing. Third-party sources indicate an annual starting point around $5,000 with custom enterprise quotes based on assessment volume and product bundle. Implementations involve I/O psychologist consultations, ATS integrations, and multi-week onboarding processes. This is enterprise software priced and sold like enterprise software.

Who Should Use Which

The Pattern and the Alternative

Harver is the third company in this blog series that follows the PE acquisition pattern. Modern Hire — the science-first interview platform — was acquired by HireVue in 2023. Pymetrics — the neuroscience gaming pioneer — was absorbed by Harver in 2022. And Harver itself is the product of seven acquisitions assembled by Rubicon Technology Partners.

The pattern repeats because it works — for the investors. PE firms buy specialized tools, bundle them into “platforms,” sell comprehensive solutions to enterprise procurement teams, and create entities that look impressive in a capabilities matrix. The tools get maintained. Customers get served. But the original innovation in each acquired company becomes a line item in someone else’s feature list, and the roadmap bends toward whatever makes the portfolio more valuable at exit.

SageScreen was built as one thing: an AI interviewing platform that conducts adaptive behavioral interviews and produces evaluations hiring teams can trust. We’re not assembling a suite. We’re not buying our way into adjacent categories. We’re not optimizing for a capabilities checkbox that makes an enterprise buyer feel like they got everything in one contract.

We’re building a product that does one thing well enough that the one thing is worth choosing on its own.

The Bottom Line

Harver is a real company with real customers solving real problems in volume hiring. The I/O psychology behind their assessments is validated. The neuroscience from Pymetrics is genuine. The reference checking works. If you’re a BPO or contact center hiring 5,000 people a quarter and you want a single platform that filters candidates before your recruiting team interviews them, Harver is a credible option.

But if you’re looking for AI that actually conducts interviews — that talks to candidates, follows up, evaluates competency, and produces evidence — you’re looking at a different category. Harver’s “interview” is a recording booth. SageScreen’s interview is a conversation with a purpose.

And if the stability of the platform you build on matters — if you want to know that the product will be the same product in two years, not an integration target — it’s worth looking at who’s building versus who’s buying.