Apply on
Availability Status
This job is expected to be in high demand and may close soon. We’ll remove this job ad once it's closed.
RESPONSIBILITIES
Develop and implement LLM-based applications tailored for in-house legal
Fine-tune and deploy large language models to enhance their performance on legal text processing tasks
Evaluate and help maintain our data assets and training/evaluation data sets
Design and build pipelines for preprocessing, annotating, and managing legal document datasets
Collaborate with legal experts to understand requirements and ensure models meet domain-specific needs
Conduct experiments and evaluate model performance to drive continuous improvements
Interface with other technical personnel or team members to finalize requirements.
Work closely with other development team members to understand moderately complex product requirements and translate them into software designs.
Successfully implement development processes, coding best practices, and code reviews for production environments.
REQUIREMENTS
Formal training in machine learning: dimensionality reduction, clustering, embeddings, and sequence classification algorithms
Experience with deep learning frameworks such as PyTorch, Tensorflow and Hugging Face Transformers.
Practical experience in Natural Language Processing methods and libraries such as spaCy, word2vec, TensorFlow, Keras, PyTorch, Flair, BERT
Practical experience with large language models, prompt engineering, fine-tuning and benchmarking, using frameworks such as LangChain and LlamaIndex
Strong Python background
Knowledge of AWS, Google Cloud Platform, Azure, or other cloud platform
Understanding of data modeling principles and complex data models.
Proficiency with relational and NoSQL databases as well as vector stores (e.g., Postgres, Elasticsearch/OpenSearch, ChromaDB)
Knowledge of Scala, Spark, Ray, or other distributed computing systems highly preferred
Knowledge of API development, containerization, and machine learning deployment highly preferred
Experience with ML Ops/AI Ops highly preferred