Beyond keyword filtering. A look at what happens when job matching is built around cognitive profiles instead of resume phrases.

Traditional job matching has a structural problem. It was designed to quickly screen resumes for keywords, reduce a large applicant pool to a small one, and get hiring managers a shortlist. The efficiency of that system is real. The cost of that efficiency, for neurodivergent candidates, is also real, and it is almost entirely invisible to the people running the system.

A candidate whose resume does not use the exact phrases in the job description gets filtered out regardless of whether they can actually do the work. A candidate who took a non-linear career path, which many neurodivergent professionals do, gets filtered out because the system was not designed to read anything other than a conventional progression.

The version of AI matching that most neurodivergent candidates encounter is just the old keyword system running on faster hardware. It is not intelligence, it is pattern matching at speed. What would actually intelligent matching look like?

Starting with the Candidate, Not the Resume

The most meaningful shift happens when the matching system is designed around the full picture of who the candidate is rather than the narrow slice their resume captures. This means incorporating self-reported information about working style, sensory preferences, communication tendencies, and cognitive strengths alongside experience and skills.

For neurodivergent candidates, this changes the mechanics of matching fundamentally. A candidate whose profile indicates strong hyperfocus, preference for deep work, and comfort with asynchronous communication can be matched to roles and teams where those working conditions are genuinely available, rather than to roles that just list the right keywords.

Reading the Employer Side More Carefully Too

The other half of matching is understanding what the employer actually needs, not what their job description says they need. Job descriptions are written for applicant volume, not accuracy. They include keywords that are there to satisfy HR rather than to describe the real work.

A matching system that reads between the lines of the job description, that incorporates information about the team's actual working patterns, the manager's communication style, and the real demands of the role, can match candidates to roles far more accurately than keyword-based systems ever could.

The Five Dimensions That Matter Most for Neurodivergent

Mentra's matching system evaluates candidates and roles across five dimensions. Overall fit looks at the holistic match between the candidate's profile and the role. Keyword alignment handles the conventional technical skills match. Behavioural compatibility evaluates whether the working style required by the role matches the way the candidate actually works. Composite scoring integrates the other dimensions into a single match score. Seniority calibrates the match against experience level.

The five-dimension approach is not complicated for its own sake. It is complicated because the real problem is complicated. Matching a neurodivergent candidate to a role that will actually work for them requires understanding more than what words appear on the resume.

Bidirectional Learning

The next generation of matching goes further. Instead of evaluating a candidate against a static role description, it learns from outcomes. When candidates are placed in roles, the system observes how the placement actually went. When interviews result in offers, the patterns that produced that outcome get reinforced. When candidates leave roles quickly, the signals that predicted poor fit get weighted more heavily in future matches.

This bidirectional learning is particularly important for neurodivergent matching because the signals that predict long-term fit for neurodivergent professionals are different from the signals that predict fit for the general workforce. A static system, no matter how well designed, will always be calibrated against the majority pattern. A learning system can develop distinct matching patterns for different cognitive profiles.

What This Means for Candidates

For a neurodivergent candidate, the practical difference is substantial. You are not just being evaluated against whether your resume contains the right phrases. You are being matched based on whether the role, the team, and the environment are actually likely to work for the way you operate.

The better the matching, the fewer interviews you have to sit through for roles that were never going to be a good fit. The better the matching, the more your first-week experience at a new job feels like it was built for you, because structurally it was.

See what a matching system built for your brain can do at mentra.com