AI Engineer/ML Engineer - Senior Developers - AI Training - Mesa, US

prolificacademicltd; USA; Remote; USD 66.000 - 110.000; Vollzeit

Benefits dieser Stelle

Up to $80/hr

Fully remote

Skills assessment gate

Zusammenfassung

Prolific Academic Ltd. seeks experienced AI/ML engineers to join a paid expert network evaluating and improving language model outputs.

This is a flexible, remote role based in the Mesa area, with hourly compensation for completed tasks and a focus on model alignment and reasoning quality. Contributors review model explanations, audit code and notebooks, and provide structured feedback to refine RLHF processes.

Qualifikationen

• BS, MS, or PhD in Computer Science, AI, Robotics, or related quantitative field with ML focus.
• Hands-on experience building, deploying, or fine-tuning ML models in production.
• Knowledge of transformers, CNNs, and RNNs with relevant optimization techniques.
• Experience with prompt engineering, RLHF, or RAG workflows.
• Ability to reason about ML proofs and audit complex model logic.
• Strong attention to detail for detecting hallucinations and biased outputs.

Aufgaben

• Review AI explanations of architectures, loss functions, and backpropagation for technical accuracy.
• Audit ML code/notebooks for correctness and efficiency.
• Provide structured human feedback to refine RLHF and improve alignment.
• Evaluate how a model reasons through chain-of-thought prompts and identify breakdowns.
• Run benchmarks across outputs against technical taxonomies and criteria.
• Identify hallucinations, bias, data contamination, and flawed reasoning.

Kenntnisse

Machine Learning in production

Transformers

RNNs/CNNs

RLHF

Prompt engineering

Code review

Ausbildung

Machine Learning degree

Tools

PyTorch

TensorFlow/Keras

Hugging Face Transformers

AWS SageMaker

LangChain

Pinecone / Milvus / Weaviate

Jobbeschreibung

Role overview

The position invites experienced AI and machine learning engineers to join a paid expert network that supports the training and evaluation of large language models. Contributors apply deep technical judgment to assess model outputs, audit code, and provide structured human feedback that improves model alignment and reasoning quality. It is a flexible, remote participant role based in the Mesa area rather than a traditional employment position, with hourly compensation for completed tasks.

Responsibilities

• Review AI-generated explanations of model architectures, loss functions, and backpropagation for technical accuracy.
• Audit ML code and notebooks, including training loops, preprocessing scripts, and evaluation routines, for correctness and efficiency.
• Supply structured human feedback used to refine RLHF processes and improve model alignment, safety, and helpfulness.
• Critically evaluate how a model reasons through chain-of-thought prompts and pinpoint where the logic breaks down.
• Run comparative benchmarks across model outputs against defined technical taxonomies and performance criteria.
• Identify hallucinations, bias, training-data contamination, and flawed mathematical reasoning in AI-generated technical content.

Requirements

• BS, MS, or PhD in Computer Science, Artificial Intelligence, Robotics, or a related quantitative field with a machine learning focus.
• Hands-on experience building, deploying, or fine-tuning machine learning models in production environments.
• Professional-level knowledge of neural network architectures such as Transformers, CNNs, and RNNs, along with relevant optimization techniques.
• Practical experience with prompt engineering, reinforcement learning from human feedback (RLHF), or retrieval-augmented generation (RAG) workflows.
• Ability to reason about mathematical proofs behind ML algorithms and audit complex model logic.
• Strong attention to detail when spotting hallucinations, biased outputs, or logical failures.

Nice to have

• Expert proficiency in PyTorch or TensorFlow/Keras.
• Advanced Python skills with NumPy, Pandas, Scikit-learn, and Hugging Face Transformers.
• Familiarity with AWS SageMaker, Google Cloud Vertex AI, Weights & Biases, or LangChain.
• Experience with vector databases such as Pinecone, Milvus, or Weaviate for RAG evaluation.

Benefits and work setup

• Hourly compensation for completed tasks, with researcher budgets reaching up to $80 per hour.
• Fully remote work with flexible hours; many tasks are short, though some require roughly an hour of uninterrupted focus.
• A brief skills assessment gates entry, after which contributors can begin accepting paid assignments.