cleandataengineering.aiDrug-Safety Agent on Databricks, drawn simply
You are here: Architecture

Drug-Safety Agent on Databricks · Architecture

The architecture in one picture

Every agent that does real work needs the same sixteen jobs done around the model: assembling what it reads, calling it, running tools, looping, remembering, checking, asking a person, and more. We call this surrounding code the harness. This page shows the sixteen jobs and how much of each this agent has.

Goes inA request and a model
What happensSixteen layers: most of the engine is a shared library; this agent adds its own rules and tools
Comes outA draft that has been checked, costed and, once the approval layer exists, decided by a person
Blue = built. Yellow = partly built. Dashed = not built. Click a box for its page.

Most of it is borrowed, on purpose. The loop, the limits, the tool contracts and the model connection come from a shared library, agent_harness, that every agent in this family uses. What this agent writes is only what is particular to drug safety: the rules, the three tools, the statistics, the output shape and the grounding check. Read the table under Sixteen layers, a table for each layer.

Why show the gaps. A tour that only shows what works teaches little. One layer is not built at all (observability). Two more, state and human approval, exist only as local stand-ins: a SQLite store and a web page without a login. The status page puts the gaps in order.

Look inside