Public articles linked to the same research event.
arXiv The work presents and deploys Coco, an agentic platform for the hardware-software co-design lifecycle, built as four layers: a datastore, a library of typed tools, agents encoding recurring analysis workflows such as iso-execution analysis, and a platform UX whose navigation state doubles as agent context, so that every number is grounded in a SQL query rather than scraped from heterogeneous files; it reports early deployment experience with TPU architects toward a reduction in time-to-simulation and time-to-insight.
The work presents and deploys Coco, an agentic platform for the hardware-software co-design lifecycle, built as four layers: a datastore, a library of typed tools, agents encoding recurring analysis workflows such as iso-execution analysis, and a platform UX whose navigation state doubles as agent context, so that every number is grounded in a SQL query rather than scraped from heterogeneous files; it reports early deployment experience with TPU architects toward a reduction in time-to-simulation and time-to-insight.
The work presents and deploys Coco, an agentic platform for the hardware-software co-design lifecycle, built as four layers: a datastore, a library of typed tools, agents encoding recurring analysis workflows such as iso-execution analysis, and a platform UX whose navigation state doubles as agent context, so that every number is grounded in a SQL query rather than scraped from heterogeneous files; it reports early deployment experience with TPU architects toward a reduction in time-to-simulation and time-to-insight.
The work presents and deploys Coco, an agentic platform for the hardware-software co-design lifecycle, built as four layers: a datastore, a library of typed tools, agents encoding recurring analysis workflows such as iso-execution analysis, and a platform UX whose navigation state doubles as agent context, so that every number is grounded in a SQL query rather than scraped from heterogeneous files; it reports early deployment experience with TPU architects toward a reduction in time-to-simulation and time-to-insight.