How Science-Backed Skill Screening Agents Scale Quality Hiring Perfectly

Prefer to watch instead of read? Catch the full webinar replay here: https://watch.getcontrast.io/register/maki-people-live-demo-how-science-backed-skill-screening-agents-accelerate-quality-hiring-at-scale

Hiring only starts working again when you stop guessing from CVs and start measuring skills directly for every single applicant.

TL;DR

• High-volume hiring is breaking under AI-generated CVs, mass-apply behavior, and manual screening that cannot keep up.
• Shiro, our skill screening agent, replaces CV triage with a 10–15 minute, branded skills experience that turns any applicant pool into a ranked shortlist.
• Behind Shiro sits a Cambridge-led science team and validated psychometric models that predict on-the-job performance more accurately and more fairly than CVs or gut feel.
• Recruiters save thousands of hours, time to hire shrinks by up to half, early turnover drops, and candidate satisfaction routinely hits 95–99 percent.
• The future of talent acquisition is not humans versus AI, it is recruiters equipped with intelligent agents that standardize decisions and let skills speak louder than CVs.

Why hiring feels so broken right now

If you have tried to apply for a job recently, you know how demoralising it can be. You create yet another account, upload your CV, retype the same information into endless forms, hit submit, and then hear nothing. Research consistently shows that a majority of candidates never receive a response from employers, even after investing significant time in applications.

On the other side of the table, recruiters are drowning. For many roles, hundreds of CVs arrive within days. Most look identical, even more so now that AI can write a polished profile and motivation letter in seconds. Internally and externally, we see application volumes for some companies multiplying since 2021, while recruiter headcount has not followed.

The result is a funnel that feels like a firehose. Recruiters are not selecting talent, they are surviving noise. They spend their days triaging, not partnering with the business. Strong candidates give up early. Mismatches slip through. Everyone feels the cost when people land in the wrong roles.

AI has removed almost all friction from applying, yet most hiring systems still rely on the same blunt tools: CV keywords, manual screens, and unstructured interviews. CVs have very low predictive power compared with structured assessments and structured interviews, especially for early career or volume roles.

This is the gap we designed Shiro to close.

Meet Shiro: the skill screening agent

Shiro is our skill screening agent. It takes every applicant for a role and, in minutes, turns that chaos into a clear, ranked shortlist based on actual skills, not guesswork.

Instead of asking candidates to slog through form fields, Shiro invites them into a short, fully branded experience on web or mobile. In around 10 to 15 minutes, candidates complete a series of immersive tasks tailored to the role and your competency framework.

Those tasks can include:

• Realistic simulations of the job, such as handling a demanding client as a junior auditor
• Short video or audio answers to situational questions
• Visual problem solving and cognitive items
• Scenario questions with multiple possible responses
• Open text questions that test reasoning and written communication

Everything from the visuals to the tone to the avatar that welcomes candidates can be aligned with your employer brand. The experience is designed to feel more like a preview of the job than a test: intuitive, contextual, and respectful of the candidate’s time.

Crucially, Shiro does this in less time than most traditional application forms take to complete.

What candidates actually experience

Imagine you are applying for a junior role at a firm like PwC. Instead of uploading a CV into a black box, you click “Apply” and are redirected to a branded experience that feels like PwC from the first second.

You are welcomed by an on-screen ambassador. Basic details are pulled directly from the ATS, so you are not forced to retype your name or email. Because the experience uses voice or video, the system checks your microphone and camera once and then you are ready.

From there, you move through a series of realistic tasks:

• You explain out loud how you would support a colleague with a demanding client.
• You solve a short visual puzzle that tests pattern recognition.
• You choose between different options for handling a tricky situation with a stakeholder.
• You type a short response about how you would improve your own productivity.

From the candidate’s perspective, it feels like a conversation about the role, not a bureaucratic checklist. At the end, Shiro collects lightweight feedback on the experience and clearly explains next steps. Across customers, we see completion rates above 90 percent and extremely strong candidate perception, with many reporting that this kind of process improves their view of the company.

What recruiters and hiring managers see

While the candidate is immersed in the experience, Shiro is working quietly in the background.

Each answer is scored against a predefined skills framework for the role, using psychometric models and AI that have been calibrated against human reviewers. For a language task, for example, Shiro can return a CEFR level alongside detailed sub-scores such as grammar, fluency, and coherence. For a role simulation, it can break performance down by dimensions like client orientation, problem solving, and communication.

Once candidates complete the experience, all of these signals are aggregated into:

• A global score that reflects overall fit
• Individual skill scores for each dimension assessed
• Clear pass or fail flags generated by rules you configure

You might decide, for instance, that you want to automatically advance candidates who score above 70 percent overall and at least B1 in English, and instantly reject those who fall below that bar. Shiro applies that logic consistently, without fatigue or bias.

Recruiters can always drill into the detail: rewatch video answers, read transcripts, and understand why a candidate scored the way they did. But in most cases, they simply operate from their ATS. Shiro integrates with more than 30 ATS providers and pushes scores, statuses, and workflow triggers directly into the system of record.

The outcome is simple: instead of 500 CVs and a vague sense of “good profiles”, recruiters open a requisition and see a ranked list of candidates with objective evidence behind every decision.

The science engine behind Shiro

Shiro is not just a collection of clever question types. It is powered by a science engine that combines modern psychometrics with AI.

Maki’s Head of Science, Dr Aiden Loe, spent his career at the Cambridge Psychometrics Centre working on models that link assessment performance to real-world outcomes. He has published extensively on the effectiveness of cognitive and behavioral assessments, and his work underpins how we design, validate, and monitor our tests.

The research is clear:

• Structured, job-specific assessments and structured interviews are among the most predictive and equitable selection methods.
• Composite behavioural measures and situational judgement tests can match the predictive power of pure ability tests, with smaller group differences and better fairness outcomes.
• CVs and unstructured interviews, which still dominate many hiring processes, sit near the bottom of the predictive ladder.

Shiro operationalises this science in three ways:

1. Multi-measure assessment
   Each experience blends cognitive, behavioural, and job-specific items so that you are not relying on a single test or proxy.
2. Standardised scoring
   For each skill, Maki’s science team and subject matter experts design scoring grids that describe what “good” looks like at different levels. Human reviewers and AI models are calibrated against these grids until AI scores align with expert judgement, then continuously monitored in production.
3. Focus on content, not appearance
   When we use AI to score open responses, we explicitly ignore facial expressions, tone, or other biometric signals. We focus solely on the content of what candidates say or write, because that is where the job-relevant skill lives.

This combination allows Shiro to deliver human-level scoring quality at a scale and speed no human team could match, while maintaining fairness, transparency, and auditability.

Built for enterprise trust and security

Enterprise HR teams cannot compromise on security, compliance, or reliability. Shiro sits within the broader Maki People platform, which is ISO 27001 certified and built on hardened cloud infrastructure, with end-to-end encryption and strict access controls.

We design and audit our systems against emerging regulations including the EU AI Act and local AI and hiring laws. We also apply practices such as:

• WCAG-aligned accessibility and reasonable adjustments for candidates
• Bias auditing and demographic analysis to monitor fairness
• Blind assessments, where requested, so that personal information does not leak into decision making

For large enterprises, Shiro plugs cleanly into the existing stack. Data flows between ATS, HRIS, and Maki are clearly mapped. Recruiters work primarily where they are most comfortable, while Shiro handles the hidden complexity in the background.

The impact: speed, quality, and experience at scale

Shiro was designed for organisations that need to hire fast and right across thousands, without lowering the bar.

Across customers, we see a consistent pattern of impact when skills replace CVs at the top of the funnel:

• Time to hire shrinks: recruiters are no longer manually screening hundreds of CVs or running repetitive phone screens. Time to shortlist collapses from days to minutes, and overall time to hire can fall by 30 to 50 percent.
• Recruiter efficiency soars: teams can absorb volume spikes and run large campaigns without adding headcount. For some clients, Shiro and our other agents have saved hundreds of thousands of recruiter hours by automating screening and first-round evaluation.
• Quality of hire improves: when every shortlisted candidate has been screened on the skills that actually drive success, interview to offer ratios improve and mis-hires drop. Several customers see early turnover reductions of 20 to 30 percent for top scorers.
• Candidate experience becomes a differentiator: instead of a black box CV upload, candidates experience a fast, fair, and contextual journey that tells them something about the role and about themselves. Satisfaction scores of 8.5–9.5 out of 10 and 95–99 percent positive perception are now common for our clients.

In practice, this looks like Wise delivering 300 customer care hires on time from tens of thousands of applications with just two recruiters, or global brands screening tens of thousands of candidates per campaign while maintaining completion and satisfaction rates close to 100 percent.

Skills over CVs: where this is heading

The rules of hiring have changed. AI has made it simple for anyone to generate a perfect, keyword-rich CV. Volume has exploded. Traditional filters no longer separate signal from noise.

Throwing more people at the problem does not work. It is expensive, it burns out teams, and it still relies on inconsistent human judgement.

Boards and executives now expect hiring to drive measurable gains in performance, retention, and customer experience, not just hit time-to-hire or cost-per-hire targets.

Shiro is part of a broader shift: from manual, CV-driven processes to intelligent, skills-based hiring engines that evaluate every candidate, predict who will succeed, and learn with every outcome.

In that world:

• Candidates feel respected and informed.
• Recruiters act as strategic partners, not gatekeepers.
• Hiring decisions become evidence-based conversations, not subjective debates.
• Talent acquisition finally has the data to prove its impact on business performance.

FAQ: Shiro, skills, and AI scoring

Can we share results and personalised feedback with candidates?
Yes. Shiro can generate structured feedback reports for candidates automatically or on demand. These can include high-level results, strengths by skill, and tailored suggestions for improvement, along with curated resources such as books, podcasts, or online courses. Organisations can decide how detailed that feedback should be and can plug in their own learning resources if they wish.

How exactly does the AI score candidates?
AI scoring in Shiro is optional but increasingly adopted because it reveals richer signals than simple right or wrong scoring. For each open response, we score only the content, not appearance or emotion. Answers are evaluated against scoring grids that describe performance levels for each skill, designed by IO psychologists and subject matter experts. Human raters and AI models are both run on the same sample of answers, then we calibrate the AI until it matches human judgement and monitor it continuously in production to ensure stability and fairness.

Does Shiro offer enterprise-grade security and compliance?
Yes. Maki People is ISO 27001 certified, uses enterprise cloud infrastructure, and designs products to align with regulations such as the EU AI Act and local AI and hiring frameworks. Data is encrypted in transit and at rest, access is tightly controlled, and we work closely with client security and legal teams during implementation.

Is this about replacing recruiters with AI?
No. Shiro is designed to free recruiters from low-value, repetitive work so they can focus on what only humans can do: building relationships, exercising judgement, and partnering with the business. The most successful teams are those that pair human expertise with intelligent agents that standardise and scale good decisions, not those that try to remove humans from the loop.

‍

See the full live demo and Q&A on demand: https://watch.getcontrast.io/register/maki-people-live-demo-how-science-backed-skill-screening-agents-accelerate-quality-hiring-at-scale

‍

‍

At Maki People, our mission is to give HR more-than-human powers by combining science-backed assessments, autonomous agents, and real outcome data into one intelligent hiring engine. Shiro is often the first step on that journey: the moment where CVs stop deciding your future, and skills finally get the signal they deserve.

If you want to see what Shiro could look like in your own hiring funnel, our team would be happy to design a tailored experience and walk you through candidate and recruiter flows end to end.