About mpathic.ai

mpathic is keeping humans safe in the AI era through automated tools and expert datasets that are rooted in psychology and powered by clinicians.

We are a Series A start-up backed by Tier 1 investors including Foundry.vc and Next Frontier Capital.

Position Overview

mpathic is seeking a Senior Clinical Safety Lead to support high-quality annotation, labeling, red teaming, and AI safety evaluation work across sensitive human data projects.

You will serve as a senior quality lead for expert-led evaluation programs, with a focus on annotation QA, labeling consistency, calibration, reviewer training, adjudication, and clinical quality standards. You will help ensure that human data outputs are rigorous, auditable, clinically grounded, and ready for use in model evaluation, benchmarking, safety testing, and customer-facing reporting.

This role is designed for a licensed or clinically trained expert with strong judgment in psychological safety, qualitative review, clinical supervision, and rubric-based evaluation. The work does not involve providing therapy, crisis intervention, diagnosis, or on-call clinical care.

This is a full-time, remote role reporting to Human Data / AI Safety leadership and partnering closely with Senior TPMs, Human Data Operations, clinical experts, research, product, engineering, and customer delivery.

Key Responsibilities

Clinical QA, Annotation & Labeling Review

• Review annotated and labeled datasets, roleplays, and model responses for clinical appropriateness, label consistency, rubric adherence, and safety alignment.
• Serve as a senior escalation point for ambiguous labels, disagreement clusters, edge cases, and sensitive conversations.
• Make and document adjudication decisions within defined project guardrails.
• Ensure evaluation outputs are rigorous, clinically grounded, internally consistent, and customer-ready.
• 
  

Reviewer Training & Calibration

• Train and onboard clinical experts, reviewers, annotators, and red teamers on rubrics, labels, taxonomies, and evaluation standards.
• Run calibration sessions to improve reviewer consistency and reduce disagreement.
• Develop training materials, gold-standard labels, anchor examples, and reviewer guidance.
• Identify reviewer drift or inconsistent label use and translate findings into updated guidance.

QA Systems, Rubrics & Taxonomies

• Build and improve QA workflows for annotation, labeling, red teaming, benchmarking, and model evaluation projects.
• Partner with behavioral science, QA, research, and operations teams to refine rubrics, labeling guides, taxonomies, and edge-case guidance.
• Maintain gold sets, severity calibration standards, and reviewer decision logs.
• Identify rubric gaps, unclear definitions, and recurring sources of disagreement.

Red Teaming & AI Safety Evaluation

• Review roleplays, model responses, safety signals, and behavioral edge cases.
• Evaluate how models respond to distress, vulnerability, crisis-related content, boundary issues, and other sensitive human interactions.
• Identify recurring model failure patterns, safety risks, and response-quality concerns.
• Translate clinical QA findings into structured feedback for reviewer guidance, product workflows, and customer reporting.

Cross-Functional Collaboration

• Partner with Senior TPMs and Human Data Operations to support fast-moving project execution without sacrificing clinical quality.
• Collaborate with clinical experts, QA leadership, product, and engineering to improve evaluation standards, review workflows, and tooling.
• Support customer delivery by ensuring findings are interpretable, trustworthy, and grounded in high-quality human review.

What You'll Accomplish

In your first 60–90 days, you'll…

• Build fluency in mpathic's annotation, labeling, red teaming, AI safety evaluation, and clinical QA workflows.
• Review active project outputs for consistency, clinical plausibility, and safety alignment.
• Train expert reviewers and run calibration sessions to improve alignment.
• Serve as an escalation point for ambiguous labels and sensitive edge cases.
• Deliver high-confidence adjudicated outputs and identify opportunities to improve QA workflows.

In your first year, you'll…

• Own clinical QA standards across multiple annotation, labeling, red teaming, and AI safety evaluation projects.
• Develop scalable reviewer training, calibration, certification, and feedback systems.
• Improve reviewer consistency, inter-rater reliability, rubric clarity, and labeling quality.
• Build gold-standard examples, clinical playbooks, and edge-case guidance.
• Shape mpathic's long-term approach to clinically grounded AI safety evaluation.

You'll Thrive in This Role If You…

Have 5+ years of experience in one or more of the following:

• Clinical supervision, training, peer review, or quality review.
• Clinical research operations, qualitative review, annotation QA, or rubric-based evaluation.
• Clinical research writing, including reports, manuscripts, and white papers.
• Trust and safety, human-centered AI, behavioral science workflows, or sensitive-content evaluation.
• AI evaluation, red teaming, model benchmarking, or human-in-the-loop review systems.

And you bring clinical expertise such as:

• Clinical licensure or advanced clinical training, such as LCSW, LMFT, PsyD, PhD, MD, or equivalent applied experience.
• Research writing experience, with evidence of peer-reviewed manuscripts in psychology, social science, and/or human-computer interaction.
• Deep familiarity with psychological stress, trauma, crisis response, risk assessment, safety planning, or mental health frameworks.
• Experience evaluating emotionally complex conversations for safety, appropriateness, boundaries, and harm risk.
• Strong judgment in ambiguous edge cases involving distress, vulnerability, or sensitive disclosures.

You are especially strong at:

• Reviewing annotations and labels for consistency, nuance, and clinical appropriateness.
• Training experts to apply nuanced rubrics, taxonomies, and labeling schemes consistently.
• Writing research reports that speak to clinical nuance and model performance.
• Running calibration sessions and making difficult adjudication decisions.
• Synthesizing qualitative review data to identify failure modes, reviewer disagreement, and rubric gaps.
• Maintaining clinical quality while moving at the pace of client delivery.

Additional Requirements

• Willingness to sign required NDAs, confidentiality agreements, and customer agreements.
• Comfortable working with sensitive mental health, safety, risk, and emotionally complex content.
• Available for recurring team meetings, calibration sessions, and occasional flexible hours based on project needs.
• Eligible to work full-time in the United States.

Compensation & Benefits

• Competitive base salary
• Equity
• Company-funded health, dental, and vision insurance
• 401(k) with 4% company match
• Well-being programs
• Flexible paid time off
• Remote-first work environment
• Mission-driven work focused on AI safety, trust, and responsible innovation

To Apply

Please submit your resume along with a brief cover letter describing your experience with clinical QA, annotation or labeling review, rubric-based evaluation, reviewer training, calibration, red teaming, AI safety, or related work.

Applications must be submitted through mpathic's official recruiting portal. Only applications submitted directly by the applicant through our portal will be considered.