RoleSage Review (2026): Pricing, Features & Honest Verdict

RoleSage is skills-based hiring software that replaces the resume with an evidence graph. Candidates build a profile where every skill claim links to the work that proves it, and hiring teams query that evidence instead of filtering keywords. Our verdict: the most thoughtful answer we have seen to AI-written applications, sold at a price that assumes you hire constantly. Best for: small teams hiring more than four roles a year. Price: from $50 per opening, or $90 per seat/mo (free plan: for candidates). Rating: 7/10.

What is RoleSage?

RoleSage is built around one observation: AI has made every applicant look excellent on paper, so the resume no longer carries signal in either direction. Candidates compress a career into two pages and never learn what the screen saw. Hiring teams get hundreds of near-identical applications written by the same tools, tighten the filters in response, and cut real talent along with the noise. More automation on each side makes the relationship worse, and both sides know it.

The product’s answer is a shared evidence layer. A candidate’s CV is parsed into structured roles, activities and outcomes, which the candidate then reviews, corrects and enriches before anyone relies on it. Every skill claim links to the specific activity where it was demonstrated, producing a graph rather than a keyword list. LinkedIn verification and optional independent professional relationship support sit alongside as separate trust signals. On the hiring side you ask questions in natural language, and answers come back grounded in the candidate’s actual work history with sources attached.

RoleSage is the work of Mark Wallis and it launched on 8 August 2026. It is a two-sided product, which is the hardest kind to build and the hardest to bootstrap, and the entire design only pays off if both sides show up.

What are RoleSage’s key features?

Evidence-backed match bands instead of a score

Applicants land in match bands with an inspectable breakdown, not a mystery percentage. You can see which evidence produced the band and follow each skill claim back to the work that supports it. This is the correct design and it is worth stating why: an unexplained 87 percent match is unusable in a hiring decision you may later have to defend, whereas “strong on distributed systems, evidenced by these three activities, thin on team leadership” is something a hiring manager can act on and a candidate can argue with.

A minimum match threshold that gates the application

Candidates see match context before they apply, and hirers set a minimum match level candidates must reach in order to apply at all. This is the actual fix for the application avalanche, and it works on the input rather than the output. Filtering 400 applications faster does not solve the problem. Not receiving 340 of them does. It also treats candidates better, since they learn where they stand before spending an evening on a cover letter.

Sage, an assistant grounded in the candidate’s evidence

You ask “has this candidate run high-traffic systems” and get an answer from their real work history with sources rather than a summary of their self-description. Around it sits AI role definition that turns a rough brief into a structured role with required skills and must-have flags, plus role analysis that flags biased language, compliance risks and anything likely to deter good candidates. The bias and compliance check on the job description is a small feature doing unglamorous work, and most ATS products still do not have it.

A complete ATS underneath

This is not a scoring widget bolted onto a spreadsheet. There is an application pipeline with shortlisting, info requests, interviews and rejections plus shared team notes; an interview hub with calendar invites, interviewer feedback capture and AI-synthesised summaries; a talent pool with follow-up dates and relationship history; funnel reports covering drop-off and source of hire; offers with approval routing, day-one handoff prep and an audit trail; group-based permissions; role templates; and context prompts that keep the AI grounded in your company. For a first release from one person, that is a lot of surface area.

Candidates own the profile

The candidate reviews the parsed profile before it represents them, corrects it, and controls what it says. There are no page limits, so case studies, portfolio links and outcomes get room a resume never had. Candidate access is free at its core. Whether that is enough to build the candidate side of a marketplace is the open question, but the incentive design is right.

How much does RoleSage cost?

RoleSage sells two ways, and both are published. Opening credits cost USD $50 for one credit or USD $200 for five, a 20 percent saving. One credit publishes one role opening, credits never expire, and they are shared across the employer workspace. The paid-seat subscription is USD $90 per paid seat per month for unlimited published openings, with fair usage applying to AI-assisted workflows. Every new company gets one free opening credit, so you can run a real hire end to end before paying anything.

Only users who publish roles and create openings need a paid seat. Everyone else on the hiring team collaborates for free, and every subscribed workspace includes at least one paid seat. Cancellation is self-serve, monthly payments are refunded in full, and annual payments are refunded for subscription months not fully completed. Credits and subscriptions are billing choices rather than feature tiers, so the entire feature list is available either way. That last point is unusual and very customer-friendly.

The maths is simple. Four openings a year on credits costs $200. A single seat for twelve months costs $1,080. The crossover sits around 21 openings a year, so almost every company hiring occasionally should stay on credits, and only teams hiring continuously should subscribe. Compared with the category this is cheap: Ashby’s Foundations plan is published at $400 a month, and Workable starts at $299 a month and runs to $599 for Standard. A company hiring four roles a year pays RoleSage $200 and Workable at least $3,588.

How risky is a two-sided hiring product from one founder?

More risky than the feature list suggests, and worth thinking about before you route real candidates through it. RoleSage only delivers its core promise when candidates have built evidence profiles. On day one, most applicants to your role will not have one, which means the parsing pipeline is doing the work from an uploaded CV and the evidence layer is only as good as what the candidate bothers to enrich. The tool degrades gracefully into a decent ATS in that case, but a decent ATS is not why you would pick it.

There is also key-person risk. This is a solo build across an unusually wide surface: ATS, interview scheduling, offers, reporting, permissions, an AI assistant and a candidate-facing product. Any one of those is a company. All of them from one person means either impressive leverage or thin coverage in places you will not discover until you are mid-hire. There are no published customers, no case studies and no security certifications on the site.

The mitigation is built into the pricing, which is why we do not treat this as disqualifying. Your first opening credit is free and credits never expire. Run one real, non-urgent hire through it start to finish, see how many candidates complete an evidence profile, and buy the five-credit pack only if that number is not depressing.

Who is RoleSage best for?

Use RoleSage if you are a startup or small company hiring a handful of roles a year, you are drowning in AI-written applications, and you care about being able to explain a hiring decision afterwards. The credits model fits that shape exactly and costs less than one month of any established ATS. It suits engineering and technical hiring particularly well, where skills really do link to identifiable pieces of work.

Use it too if fairness and auditability are load-bearing for you, whether because of regulation, an internal equity commitment or a previous bad experience with a black-box screening vendor. Transparent criteria, inspectable reasoning and human decision-making at every step are the explicit design principles here, not a compliance page.

Skip RoleSage if you are hiring at volume for roles where evidence profiles will not exist, such as high-turnover retail or hospitality, because the candidate side will not be populated and you are left with a young ATS. Skip it if you need deep integrations with your HRIS, payroll and background check vendors, which is where Ashby and Workable have years of work you cannot replicate. And skip it if you cannot be an early customer on something as consequential as hiring.

Best RoleSage alternatives

Ashby is the strongest all-in-one recruiting and analytics product for growth-stage companies, with the reporting depth RoleSage will not match for years. Its Foundations plan is published at $400 a month, with Plus and Enterprise above it. Pick Ashby if you hire continuously and analytics drive your process; the price only makes sense at real volume.

Workable is the practical mid-market choice, with job board distribution, a large integration catalogue and an AI sourcing agent sold as an add-on that runs on credits. Starter is $299 a month and Standard is $599, with a 15-day trial of the full Standard feature set. Better distribution than RoleSage, nothing comparable on the evidence layer.

Your existing ATS plus a structured work-sample exercise is the unglamorous alternative and it addresses the same problem. If AI-polished applications are the issue, a real work sample separates candidates more reliably than any parser. RoleSage is trying to get that signal without the extra unpaid work you are asking candidates to do, which is a better goal, but the old method still works.

Final verdict: is RoleSage worth it?

RoleSage has diagnosed the problem more precisely than anything else we have reviewed in hiring this year. Resumes stopped carrying signal, filtering harder makes it worse, and the answer is to change what candidates submit rather than how fast you reject it. Grounding every skill claim in a specific piece of work, letting hirers interrogate that evidence, and gating applications on a match threshold rather than sorting the flood afterwards are three good ideas that fit together. The pricing is honest, the free first opening is a real trial, and refusing to split features across tiers is the right call.

The gap between the idea and the reality is the candidate side, and no amount of product design closes it on launch day. Until enough candidates carry evidence profiles, you are buying a competent young ATS with an unusually good roadmap, from one founder, with no published references. That is a reasonable bet at $50 for your next opening and a poor one at $1,080 a year sight unseen. Take the free credit, run one hire, and let the completion rate on those evidence profiles make the decision for you.