Skip to content

Screening and assessments

Screening and assessments

Get to know what candidates can do.

Combine résumé evidence, conversational AI screening and practical assessments to understand each candidate. Elevator by Superset gives your team a fuller basis for deciding who to meet next.

Request a demo See the assessment domains

TalentLens

See why a candidate fits the role.

TalentLens assesses résumé evidence against six weighted criteria and the requirements of your requisition. Explanations accompany the scores, giving recruiters context to review with the hiring manager.

What’s included
  • Three of the six axes are arithmetic, not model output, which is why the number survives being questioned
  • No campus signals: no CGPA, no institution tier, no internships
  • The same fit logic the careers site shows candidates, so nobody sees “strong match” where the recruiter sees a D
How AI is governed in Elevator

AVA screening

Make the first conversation more accessible.

AVA runs an asynchronous conversational screen within your pipeline. Candidates respond in their own time; recruiters review competency summaries, communication scores and integrity checks before confirming the next step.

What’s included
  • Five competencies for operations roles: scenario judgement, customer empathy, process adherence, communication, role knowledge
  • Recommended action is advance, review or reject; a recruiter confirms every one
  • Built for the high-volume operations template: assessment first, AVA screen, then an operations round
See the pipeline templates

Assessment batteries

Match the assessment to the work.

Choose role-specific assessment batteries across 12 domains, including software engineering, data and enterprise systems. Place them at the appropriate point in your pipeline for the role and seniority you are hiring.

What’s included
  • Software engineering covers core languages, backend, frontend and mobile
  • ERP and enterprise systems covers the platforms your teams actually run
  • Product and architecture covers system design and technical leadership

Formats and integrity

Give skills more ways to show up.

Use 24+ formats, from coding and practical exercises to Skill Studio work simulations and Eloquence® communication assessments. Superset’s proctoring infrastructure supports the assessment experience, with integrations for 30+ external providers.

What’s included
  • Capacity for 800,000 concurrent assessments
  • Skill Studio work simulations and Eloquence® English communication assessment
  • 30+ assessment-provider integrations, with results landing on the candidate record
Explore integrations

How AI is used

AI drafts. Rules route. People decide.

TalentLens and AVA produce scores and recommendations; they never commit a decision. A written, versioned rule decides what happens at a given score, and a recruiter’s own action advances or rejects a candidate. Candidate data is never used to train a model. The full map of every model in the product and what it is allowed to do is on the AI in Elevator page.

What is included

In screening and assessments

  • TalentLens résumé scoring on six weighted competency axes, weights summing to 100
  • Written explanation on every score, quoting the evidence it used
  • Rubric chosen by title, scored against the requisition’s own band, grade and target date
  • No campus signals: no CGPA, no institution tier, no internships
  • AVA asynchronous AI screening as a pipeline stage, on the candidate’s own time
  • AVA summary card with per-competency signal, communication score and integrity checks
  • Recommended action (advance, review, reject) that a recruiter confirms
  • Role-specific assessment batteries across 12 domains, pitched at the grade
  • 24+ assessment formats, from live coding to architecture whiteboard simulation
  • Skill Studio work simulations and Eloquence® English communication assessment
  • Proctoring and integrity from the stack that runs 12 million assessments a year
  • 30+ external assessment providers integrate, with results landing on the candidate record

Explore the rest of the platform

Screening and assessments, answered

What is AI résumé screening?
AI résumé screening is the use of a model to read résumés and rank candidates against a role, so a recruiter reviews a shortlist instead of a pile. In Elevator the model is TalentLens: it scores six weighted competency axes against the requisition’s own band, grade and target date, and writes an explanation for every score that quotes the lines it relied on. Three of the six axes are arithmetic rather than model output, which is why the number stands up when a hiring manager questions it.
What is an asynchronous AI interview?
An asynchronous AI interview is a screening conversation a candidate completes on their own time, with an AI interviewer asking the questions rather than a recruiter on a call. Elevator’s AVA runs it as a pipeline stage after the skill assessment. The candidate’s answers arrive as a summary card with a signal per competency, a communication score and integrity checks, ready for a recruiter to act on.
Does the AI make hiring decisions?
No. TalentLens and AVA produce scores and a recommended action; a recruiter confirms or overrides, and their name goes on the decision. What happens at a given score is a written, versioned automation rule set by your recruiting operations team, not the model’s judgement. The same rule fires the same way every time and is reconstructable from the audit log.
Which assessment formats are available?
Elevator offers 24+ formats, including live coding, timed MCQs, case-based evaluation, real-world debugging, architecture whiteboard simulation, practical sandbox execution and AI-evaluated structured interviews. Skill Studio adds work simulations that evaluate what a person does rather than what a résumé claims, and Eloquence® assesses English communication. Formats are combined into role-specific batteries per domain and grade.
Can we use our existing assessment provider?
Yes. 30+ external assessment providers integrate with Elevator, and their results land on the candidate record alongside Superset’s own assessments. A provider assessment can gate a pipeline stage the same way a native one does, so automation rules and scorecards treat both alike. Which providers are configured is agreed during onboarding.
What does TalentLens score, and what does it ignore?
Elevator scores experience, not academic signals. TalentLens weighs domain depth, scope and seniority, role continuity, compensation fit, availability and evidence provenance, and ignores CGPA, institution tier and internships. Assessment batteries are role-specific across 12 domains and pitched at the grade, so a senior backend engineer sits a different test from a junior one.
How is integrity protected during assessments?
Every assessment runs on the same proctoring and integrity stack that delivers 12 million assessments a year on Superset, with capacity for 800,000 concurrent sessions. Integrity checks also run inside the AVA screen and surface on the summary card. Identity and experience fraud detection continues through interviews and verification, and is covered on the offers page.

Your next team starts here

Your hiring.
All together.

A walkthrough of Elevator, shaped around your roles, your people and your process.

Request a demo Explore the platform