Senior Machine Learning Engineer
Current• Created sophisticated data pipelines using Databricks and PySpark as the engineer responsible for dataset preparation in a major LLM training contract. The resulting models would be the main ones used by Chegg for question answering for some time.• Pioneered research into multimodal question answering for diagrams, creating experimental vision LLMs with HuggingFace and PyTorch toolchains.• Developed a powerful LLM testing harness for experimenting with techniques to improve llm reasoning and created a toolset for text and structured data processing that became the standard used across teams.• Ensured quality by: designing and reviewing informative and reliable data annotation tasks; performing data analyses ad hoc with scientific python; and contributing to resilient container orchestration.• Engineered features for diverse machine learning toolsets large language models, from autogluon to advanced mathematical tools of SciPy.• Promoted agile methodology in research featuring JIRA for documentation and planning, and CI/CD tools for automated testing and code-reviews.
