Hyrr Platform: The Skill-Centric Architecture

The Hyrr platform is built from the ground up to fundamentally shift the recruitment paradigm from resume-based screening to an objective, skill-centric assessment model. Every layer of the platform—from job creation and candidate application to interview processing and talent discovery—is engineered to prioritize verifiable competencies over traditional credentials.

This document details how the skill-centric architecture is implemented across the Hyrr ecosystem.

1. Defining the Requirements: Job Creation & The Two-Tier Skill Set

The skill-centric journey begins when an employer creates a job.

Instead of accepting a static block of text, the job creation process is highly structured to extract and define the exact competencies required.

• AI-Powered Description Generation: The system utilizes advanced AI to parse unstructured prompts and automatically generate structured requirements and nice-to-have criteria.
• Competency Matrices & Assessment Rubrics: When a job is submitted, it is saved with an underlying assessment rubric, a detailed competencies matrix, and a strict passing threshold.
• Two-Tier Skill Set: Our architecture categorizes skills into a two-tier system to ensure comprehensive evaluation:
  • Tier 1 (Job-Specific Verification): Skills verified dynamically during standard job interviews tailored to a specific role.
  • Tier 2 (Skill Certifications): Standalone skill certifications (e.g., "Python Developer") that evaluate specific pillars and map to standardized proficiency levels.
• Semantic Job Matching: As jobs are defined, their specific skill profiles are mathematically mapped, so they can be matched against candidate skill profiles later in the pipeline.

2. Skills Over Credentials: The Application Process

The application process measures every applicant against the job's skills from the first step, before the live evaluation.

• Résumé Screening on the Same Rubric: Résumés are screened against the same rubric and pass bar as the interview, so the first filter and the interview measure the same skills.
• Seamless Progression: Applications are reviewed via the employer portal. Moving a candidate forward to the interview phase instantly triggers an automated invitation, granting the candidate immediate access to the live AI interview.

3. Objective Evaluation: Anchoring with BARS

The core evaluation engine resides in our autonomous AI proxy and the subsequent processing pipeline, which objectively evaluates every candidate.

• Rubric-Driven Assessment: Once the live video interview concludes, the processing engine activates. It retrieves the specific competencies matrix and assessment rubric defined during the job creation phase.
• Behaviorally Anchored Rating Scales (BARS): To ensure fairness and precision, the AI evaluates the candidate's transcript using Behaviorally Anchored Rating Scales (BARS). This method anchors scoring directly into each specific skill by matching the candidate's responses against predefined behavioral examples of varying performance levels.
• Generative AI Scoring: Using these BARS-driven rubrics, the platform assesses technical strengths and behavioral strengths, ultimately extracting an exact, objective proficiency mapping for every competency evaluated.

4. The Verifiable Skill Database

The culmination of the skill-centric architecture is the Verifiable Skill Database—a talent pool searchable by the skills candidates demonstrated in interviews, where claimed skills are backed by evidence from the interview recording.

When the interview assessment concludes, the skill verification logic executes:

• Threshold Gating: The system evaluates the proficiency results. Only skills where the candidate scored above the passing threshold are deemed officially "verified."
• Evidence Attachment: For every verified skill, the system binds critical contextual metadata: a summary of exactly how it was demonstrated, alongside the specific video snippet from the interview itself.
• Semantic Talent Discovery: Because skills are stored as semantic concepts rather than rigid text keywords, employers can easily and intuitively discover talent. When an employer searches for capabilities like "Frontend Architecture," the system understands the underlying meaning and returns candidates who demonstrated those skills in a recorded interview.
• Proof of Action: The employer interface directly displays the written verification summary alongside an embedded video player queued to the exact moment where the candidate demonstrated the searched capability.