Conduct independent applied research on context compression for LLM pipelines. Train machine learning models and transformers, investigate research hypotheses, contribute to customer-facing solutions, and work on fast-paced, time-sensitive problems with limited supervision. The internship lasts three months, with a potential full-time offer based on performance.

Build and own full-stack infrastructure for an AI context-compression platform. Responsibilities include developing backend services, working with concurrency and distributed systems, deploying and scaling production systems on AWS, contributing across Python and frontend technologies, and operating in a fast-paced research-driven startup environment.