01 / 2026
Selected WorkAgent Runtime Systems
A case study on designing reliable agent execution flows for production-grade AI systems.
Read case study →[SYSTEMS]
Case studies organized by problem category: agents, retrieval, conversation, distributed product systems, ML optimization, and efficient AI experiments.
01 / 2026
Selected WorkA case study on designing reliable agent execution flows for production-grade AI systems.
Read case study →02 / 2026
Selected WorkDesigning retrieval flows that balance latency, provenance, freshness, and answer quality.
Read case study →03 / 2025
Selected WorkA systems view of conversation as state, policy, memory, and interface design.
Read case study →04 / 2025
ArchiveBuilding product infrastructure that survives growth, dependency failure, and changing product direction.
Read case study →05 / 2026
ExplorationA case study on making ML iteration tighter, cheaper, and easier to evaluate.
Read case study →06 / 2026
Research AdjacentA collection of experiments around token efficiency, compact context, and smaller model roles.
Read case study →