Build and own production applied-AI systems focused on information retrieval and LLM pipelines. Design experiment frameworks and evaluation pipelines, optimize for metrics (precision, recall, F1), run benchmarks, perform failure analysis, and iterate to improve real-world system performance at scale.

Design, build, and operate large-scale distributed production systems focused on latency, reliability, and correctness. Own systems end-to-end, reason about architecture and failure modes, optimize performance vs. correctness, and ship production-quality infrastructure rapidly in an early-stage startup environment.