AI Engineer

Join Crossing Hurdles as an AI Engineer, working remotely on cutting-edge AI model development. This hourly contract position offers a flexible part-time commitment of 20-30 hours per week. Responsibilities include building and optimizing NLP pipelines for text classification, tokenization, entity extraction, and RAG workflows, as well as designing scalable text analytics for multilingual datasets.

You will implement and fine-tune NLP models using Hugging Face Transformers, spaCy, and PyTorch, developing robust evaluation frameworks for accuracy and performance. Constructing retrieval and neural search components that integrate NLP with IR systems is also key. This role involves working with JSON and JSONL datasets to create efficient, production-ready data flows and collaborating with AI researchers to integrate NLP solutions into larger AI systems.

Ideal candidates possess a background in computer science, computational linguistics, or applied machine learning. Proficiency in Hugging Face Transformers, PyTorch, spaCy, and modern tokenization libraries is essential. Experience with multilingual corpora, text-based datasets (JSON/JSONL), and a strong understanding of NLP tasks like classification, NER, and embedding generation are required. You should have experience building scalable text preprocessing and data pipelines, with the ability to fine-tune and evaluate models across diverse benchmarks and languages. Curiosity for RAG systems and neural search integration is highly valued.

Key domains of expertise include Natural Language Processing (Tokenization, embeddings, NER, multilingual NLP), Machine Learning (Model fine-tuning, PyTorch workflows, Transformer architectures), Information Retrieval (RAG, neural search, retrieval pipelines), Text Analytics (Preprocessing, evaluation frameworks, large-scale text pipelines), and AI Systems (Integration of NLP components).