LLMs Automate SDR Workflows
Austin Hughes, CEO and co-founder of Unify, on the death of the SDR
The real shift is that LLMs are no longer just writing sales copy, they are taking over the repetitive judgment calls that used to force a human into the loop. In Unify’s case, that means reading inbound and intent signals, checking company and buyer context, deciding whether an account is worth working, and then triggering the next step automatically. The product is replacing 5 to 10 minutes of tab hopping across LinkedIn, Salesforce, Apollo, and sequencing tools with a machine decision that is good enough to run at scale.
-
The job being automated is concrete. A rep sees a website visit or inbound lead, opens LinkedIn, checks what the company sells, whether it has pricing, who the likely buyer is, and whether the account is already in CRM. Unify says those checks can now be done inside the workflow, not by a person.
-
This is why warm outbound is different from older sales automation. Outreach and Salesloft sped up sending, but still relied on humans to decide who to contact. Unify’s wedge is moving the judgment step upstream, using first party data, CRM records, and intent signals to decide whether outreach should happen at all.
-
The same pattern is showing up across the stack. Default uses embedded AI for lead scoring, research summaries, and routing, while Apollo is adding signals and workflow products as cold contact data gets commodified. The control point is shifting from owning a database to owning the workflow logic around revenue events.
From here, the winners are likely to be the tools that own both the data model and the action layer. As more qualification, routing, and research steps become machine runnable, sales software moves closer to an operating system for revenue teams, with fewer point solutions and fewer junior reps doing manual triage.