AI/ML Engineer - Model Dev & Data Pipeline
Description
AI/ML Engineer - Model Dev & Data Pipeline
Type: Temporary contract role with the potential to become a full-time position.
Source: Direct individual contract only; agencies, firm uplift rates not accepted
Location: Remote (US time zones)
Contract/Hourly: DOE
Experience: 7+ years
Contract/Hourly: DOE
Experience: 7+ years
Dates of Engagement: Oct 1 to Dec 15, 2025
Hours Per Week Available: 25 hrs (min) > to 35 hrs (max) per week
Hours Per Week Available: 25 hrs (min) > to 35 hrs (max) per week
Start Date: ASAP
Rate: Projected: $65-75/ph
Rate: Projected: $65-75/ph
About The Role
Join our client’s AI/ML team to build and deploy state-of-the-art machine learning models that power intelligent experiences for millions of users. You'll work on everything from training custom models to optimizing inference pipelines, while collaborating with world-class researchers and engineers.
What You'll Build
Model Development & Research
- Design and train custom neural networks for human-centered AI using PyTorch/TensorFlow
- Fine-tune and adapt large language models (Llama, Claude, GPT) for domain-specific tasks
- Implement novel architectures from recent papers (attention mechanisms, retrieval augmentation)
- Conduct rigorous A/B testing and evaluation of model performance in production
Production ML Systems
- Build and maintain scalable ML pipelines processing 10M+ daily inferences
- Implement real-time model serving with <100ms p95 latency requirements
- Design and optimize vector similarity search systems for multi-modal embeddings
- Create robust data ingestion and feature engineering pipelines
MLOps & Infrastructure
- Establish MLflow/Weights & Biases workflows for experiment tracking and model versioning
- Build automated training and deployment pipelines using Kubernetes and Docker
- Implement model monitoring, drift detection, and automated retraining systems
- Optimize GPU utilization and cost efficiency for training and inference workloads
Requirements
Technical Expertise
- 7+ years of hands-on ML experience with production model deployment
- Expert-level proficiency in Python and ML frameworks (PyTorch/TensorFlow/JAX)
- Deep understanding of transformer architectures, attention mechanisms, and modern NLP
- Experience with large-scale distributed training (multi-GPU, model/data parallelism)
- Strong background in statistics, linear algebra, and optimization theory
Production ML Skills
- Experience with MLOps tools: MLflow, Weights & Biases, Kubeflow, or similar platforms
- Proficiency with cloud ML services (AWS SageMaker, GCP Vertex AI, Azure ML)
- Docker and Kubernetes experience for containerized ML workloads
- Knowledge of model serving frameworks (TorchServe, TensorFlow Serving, TritonServer)
AI/LLM Specialization
- Hands-on experience with LLM fine-tuning, RLHF, and prompt engineering
- Understanding of retrieval-augmented generation (RAG) and vector databases
- Experience with multimodal models (vision-language, audio processing)
- Knowledge of model compression techniques (quantization, distillation, pruning)
Preferred Experience
- PhD in ML/AI, Computer Science, or equivalent industry experience
- Publications in top-tier conferences (NeurIPS, ICML, ICLR, EMNLP)
- Experience at AI-first companies or research labs (OpenAI, Anthropic, DeepMind, etc.)
- Contributions to open-source ML projects with significant community adoption
- Experience with edge deployment and mobile ML optimization
Tech Stack
- ML Frameworks: PyTorch, Transformers (Hugging Face), JAX, TensorFlow
- Data & Compute: Ray, Dask, Apache Spark, CUDA, Triton
- MLOps: MLflow, Weights & Biases, DVC, Feast, Great Expectations
- Infrastructure: Kubernetes, Docker, AWS/GCP, Terraform
- Databases: PostgreSQL, Redis, Pinecone, Weaviate, ClickHouse