Remote job
AI Engineer/ML Engineer - Senior Developers - AI Training - Oklahoma City, US
Job details
About this role
Role overview Work as a senior AI/ML engineer within an expert contributor network supporting research teams that train and evaluate large language models. The position is remote and project-based, based around Oklahoma City, US, and focuses on applying deep technical expertise to shape model behaviour.
Responsibilities
- Evaluate AI-generated explanations of architectures, loss functions, and backpropagation against ground-truth ML concepts. - Audit production-grade ML code, including training loops, preprocessing pipelines, and evaluation notebooks. - Provide structured human feedback to strengthen RLHF alignment workflows. - Stress-test chain-of-thought reasoning and pinpoint where models produce hallucinations, biases, or logical failures. - Design and run comparative benchmarks aligned to specific technical taxonomies and performance criteria.
Requirements
- BS, MS, or PhD in Computer Science, Artificial Intelligence, Robotics, or a related quantitative discipline with an ML focus. - Demonstrated experience building, deploying, or fine-tuning ML models in production settings. - Working command of modern neural architectures (Transformers, CNNs, RNNs) and their optimization techniques. - Hands-on experience with LLM-specific workflows such as prompt engineering, RLHF, or RAG. - Capacity to audit complex model logic, identify training data contamination, and reason about algorithmic proofs. - Meticulous attention to detail for catching subtle errors and biased outputs.
Nice to have
- Strong proficiency in PyTorch or TensorFlow/Keras, plus advanced Python (NumPy, Pandas, Scikit-learn) and Hugging Face Transformers. - Experience with cloud ML services (AWS SageMaker, Google Vertex AI) and tooling such as Weights & Biases, LangChain, Pinecone, Milvus, or Weaviate.
Benefits and work setup
- Remote, flexible contributor engagement with hourly compensation typically reaching up to $80 per hour for this skill profile. - Tasks are usually scoped to about an hour of focused work, with many running shorter.