RLHF (Reinforcement Learning from Human Feedback) is the process of training AI systems using expert human judgement. In clinical RLHF, healthcare professionals review, rate, and correct AI-generated medical content to make AI safer and more accurate.

Who can take this training?

Our training is designed for UK-registered healthcare professionals — doctors, nurses, pharmacists, researchers, academics, and allied health professionals. You need clinical expertise but no technical or AI background.

Do I need medical qualifications?

Yes. Our training is specifically for registered healthcare professionals. Your clinical knowledge is what makes you valuable as an RLHF evaluator — you bring the domain expertise that AI companies need.

How long does training take?

Foundation Training takes about 45 minutes. Phase 1 calibration (Quick Screen) is around 20 minutes for 11 tasks. Phase 2 calibration (Full Calibration) runs 60–90 minutes across the eight task types. You can pace it across multiple sessions. Optional specialist deep-dives are coming later and won't be required for certification.

Is the training free?

Yes, all training on EnterTheLoop is completely free. We invest in your training because better-trained evaluators produce higher-quality RLHF work, which benefits everyone — candidates, employers, and patients.

Training Academy

Free RLHF training, built for clinicians.

Start with a foundation course, prove your judgement through a two-phase calibration, and finish with a reliability score on record — then get matched to paid clinical AI evaluation work. No technical background needed, and completely free.

Start training — free The methodology

The pathway

How training works.

One route, four stages — paced across as many sessions as you like. A foundation course, a two-phase calibration, and a reliability score on record, then we match you to paid clinical AI work. No AI background required; the same pathway runs for every profession.

01Foundation

Complete Foundation training

RLHF fundamentals, evaluation methodology, and clinical AI safety. Everyone starts here.

~45 min

02Phase 1

Pass the Quick Screen

A short calibration that confirms your judgement is aligned with expert consensus before the full assessment.

11 tasks · ~20 min

03Phase 2

Complete the Full Calibration

A statistical calibration across all eight task types — producing your reliability score with confidence intervals.

60–90 min

04Certified

Get certified & matched

Your reliability score goes on record, and we match you to paid clinical AI evaluation roles.

Outcome

Foundation course

Everyone starts here.

The single course every evaluator takes before calibration — RLHF concepts, evaluation methodology, and clinical AI safety principles.

RLHF Foundation

Essential training for all RLHF work. Covers AI fundamentals, task interfaces, quality standards, and justification writing.

180 min7 modulesFind out more

Calibration

Two-phase calibration.

After Foundation you take a structured statistical assessment in two phases — a short screen, then the full calibration that produces your reliability score.

Phase 1~20 min · 11 tasks

Quick Screen

A short screen that confirms your judgement is aligned with expert consensus before you commit to the full calibration. Sequential stopping means you can pass early when your agreement with gold standards is consistently strong.

Sequential stopping with anti-luck guardrails
Per-item agreement thresholds against expert gold tasks
Cool-down on retry to prevent over-fitting

Find out more

Phase 260–90 min · 8 task types

Full Calibration

A complete statistical calibration across all eight task types: rating, comparison, ranking, rubric, correction, annotation, justification, and red-team. Produces the reliability score AI companies see when matching you to roles.

Coverage across all eight task types
Beta-Binomial / bootstrap confidence intervals per metric
Reliability Report (PDF + JSON) you can share with employers

Find out more

Specialist courses

Deep-dives, coming soon.

Optional deep-dives into specific task types — pairwise ranking, correction, red-teaming, annotation, and more. Not required for certification; launching once the Phase 1 + Phase 2 pathway is fully bedded in.

Coming soonTopic roster

Pairwise Comparison & Ranking
Rating & Rubric Design
Response Correction & Rewriting
Red-Teaming & Safety
Dataset Annotation
Evidence Verification & Citation Grounding
Model Evaluation & Benchmarking
Adjudication & Disagreement Resolution
Reasoning-Trace Evaluation
Tool-Use & Agentic Labeling
Multi-Turn Conversation Quality

Find out more

Stay in the loop

New clinical AI roles,
straight to your inbox.

Join the waitlist and we'll email you when roles matching your expertise open up — plus rate updates and priority matching the moment you register.

Early access to new AI roles
Weekly pay rate updates
Priority matching when you register

Common questions

Training FAQ

Common questions about RLHF training, who can take it, and how it leads to paid AI evaluation roles.

entertheloopClinicians powering AI alignment, training & safety.

Verified against

GMCNMCGPhCHCPC

entertheloop

Clinicians powering AI alignment, training & safety.

Free RLHF training, built for clinicians.

How training works.

Complete Foundation training

Pass the Quick Screen

Complete the Full Calibration

Get certified & matched

Everyone starts here.

RLHF Foundation

Two-phase calibration.

Quick Screen

Full Calibration

Deep-dives, coming soon.

New clinical AI roles, straight to your inbox.

Training FAQ

Free RLHF training, built for clinicians.

How training works.

Complete Foundation training

Pass the Quick Screen

Complete the Full Calibration

Get certified & matched

Everyone starts here.

RLHF Foundation

Two-phase calibration.

Quick Screen

Full Calibration

Deep-dives, coming soon.

New clinical AI roles, straight to your inbox.

Training FAQ

New clinical AI roles,
straight to your inbox.

New clinical AI roles,
straight to your inbox.