We are building production language systems that extract structured intelligence from messy, ambiguous data at scale. As a Mid-Level NLP Engineer, you will own the pipeline from raw text ingestion to production inference, training and deploying models that handle millions of queries daily. This is a fully remote role for an engineer who wants to solve hard representation and latency problems in production, not just run experiments in notebooks.
Responsibilities
Train, fine-tune, and evaluate transformer-based architectures for information extraction, named entity recognition, and semantic search.
Optimize inference pipelines for sub-50ms latency using tools like ONNX Runtime, TensorRT-LLM, and vLLM.
Build automated evaluation harnesses and data labeling pipelines to detect model drift and regressions.
Architect and maintain retrieval-augmented generation (RAG) systems powered by vector databases.
Ship clean, tested, containerized code into distributed Kubernetes environments.
Requirements
3+ years of experience building and deploying machine learning systems, with at least 2 years focused specifically on NLP.
Production proficiency with Python, PyTorch, and the Hugging Face ecosystem.
Hands-on experience fine-tuning open-weights models (e.g., Llama, Mistral) and deploying task-specific encoders.
Solid understanding of vector search architectures (e.g., Qdrant, Milvus, pgvector) and embedding fine-tuning.
Strong engineering fundamentals: API design, Docker, CI/CD, and asynchronous programming.
Residing within +/- 3 hours of US Eastern Time to ensure reasonable collaboration overlap.
Nice-to-Haves
Experience implementing quantization (AWQ, GPTQ) and model distillation to run inference on constrained hardware.
Familiarity with distributed training frameworks such as DeepSpeed or Ray.
Contributions to open-source NLP tooling or published research at relevant conferences (ACL, EMNLP).
What We Offer
Base salary range of $140,000–$175,000 plus meaningful equity.
Dedicated access to high-performance GPU clusters for training and experimentation.
Fully remote setup with a $3,000 home office and hardware stipend.
100% employer-covered health, dental, and vision insurance for you and your dependents.
Flexible paid time off with a mandatory 3-week annual minimum.