Integrations / Baseloop / Custom AI Agent

Custom AI Agent

AI agent for content creation and web research on your table data

What this does

The most-used action across every production GTM workflow we audit. Teams chain Custom AI Agent for company qualification (ICP fit + segment scoring), reply categorization at scale (Sales Response, Disqualified, Out-of-Office, Positive Cold Reply — one team classifies 8,500+ Smartlead replies a month this way), out-of-office address extraction from reply bodies, call-transcript analysis, sanctions screening, and any one-off classification a workflow needs. Define a prompt and structured output fields once; every row gets a typed answer in seconds.

What this returns

A structured response generated by the LLM and written back to the row, formatted to the output fields you define when configuring the action.

How teams use this

Teams typically chain custom ai agent after an enrichment or scoring step: an inbound lead lands on the table, an enrichment action fills in the missing fields, a qualification step decides whether the lead is worth acting on, and custom ai agent runs only on the rows that pass. Because Baseloop maps each action's inputs to columns on your table, the same workflow can run on one record or on ten thousand without changing a thing.