AI Employer Intelligence
Labelbox distinguishes Horizon training environments and expert data from Recursion agents for recurring engineering and research work.
Matrix Employer Score™
81.87/ 100
New AI estimates using the existing scoring weights. These values have not been human-reviewed or independently verified, and small differences do not establish a reliable ranking.
Research and RL-environment roles offer demanding technical work and disclose US pay ranges. High-intensity role descriptions inform the work-life risk estimate, not measured hours.
Assessed on . Some source observations are historical; policies vary by role and location.
80
AI estimate · Medium confidence
Public-source observations inform an AI judgment, not a measurement of employee outcomes. A retrieval date does not mean the underlying observation is current.
81
AI estimate · Medium confidence
Public-source observations inform an AI judgment, not a measurement of employee outcomes. A retrieval date does not mean the underlying observation is current.
78
AI estimate · Low confidence
Public-source observations inform an AI judgment, not a measurement of employee outcomes. A retrieval date does not mean the underlying observation is current.
86
AI estimate · Medium confidence
Public-source observations inform an AI judgment, not a measurement of employee outcomes. A retrieval date does not mean the underlying observation is current.
90
AI estimate · Low confidence
Public-source observations inform an AI judgment, not a measurement of employee outcomes. A retrieval date does not mean the underlying observation is current.
90
AI estimate · Low confidence
Public-source observations inform an AI judgment, not a measurement of employee outcomes. A retrieval date does not mean the underlying observation is current.
80
AI estimate · Low confidence
Public-source observations inform an AI judgment, not a measurement of employee outcomes. A retrieval date does not mean the underlying observation is current.
60
AI estimate · Low confidence
Public-source observations inform an AI judgment, not a measurement of employee outcomes. A retrieval date does not mean the underlying observation is current.
Career Upside, Talent Density and Hiring Opportunity remain unassessed and are excluded from this total, rather than treated as zero. Workplace values estimate opportunities and risks from public information; actual experiences may differ.
These observations describe the available evidence. They are not component scores or verified employee-experience findings.
Named Dialpad case describes production NLP/LLM data-labeling workflows and improved quality with lower cost.
Same page uses 41% reduction in one summary and 48% in a quoted per-point comparison; do not collapse them into one verified percentage.
Read the sourceCompany describes Horizon RL environments, graders and post-training infrastructure.
Technical product description only; independent benchmark, paper or inspectable implementation still required for leadership assessment.
Read the sourceStaff ML Engineer role details Python, RL/SFT post-training, distributed infrastructure and agent evaluation; lists Kubernetes and common cloud/data tooling.
Specific employee-track role; do not mix with contractor recruitment or expert-network member counts. Requirements do not guarantee career outcomes.
Read the sourceObserved October 20, 2026: Labelbox distinguishes Horizon training environments and expert data from Recursion agents for recurring engineering and research work.
Company-reported scope, not verified adoption, technical superiority or employer quality. Role research prompt (analysis): Ask which product and employing arrangement the role belongs to; employee, expert-network and contractor work are distinct.
Read the sourceObserved October 20, 2026: The careers page mixes technical employee-track roles with explicitly labeled contractor opportunities, including expert recruitment.
Do not combine employee vacancies, contractor work and expert-network membership into a single hiring total.
Read the sourceSources were reviewed on 2026-10-18. Role listings may change; confirm current openings and eligibility with the employer. This profile is not a complete hiring snapshot. Unconfirmed employee counts, financial figures, visa policies and workplace ratings are omitted.
AI-generated research based on linked public sources. It may contain errors and has not been human-reviewed. No Evidence-Checked or Human-Reviewed designation is applied.
