Digital Worker

The older version of the idea came from robotic process automation, where scripted bots copied data between systems and were sometimes counted as “digital workers” on an org chart. The current version uses AI agents that can read a request, decide what to do and call tools. A typical example is an accounts-payable agent that reads incoming invoices, matches them to purchase orders and flags mismatches for a person to review.

The framing is deliberate. Vendors such as Salesforce describe digital labor as agentic AI doing operations that conventionally needed human judgment, partly because it lets buyers compare an agent’s cost to a salary. That comparison is a sales argument, not a measured result.

Guests use the framing for how agents should be managed. In episode 60, at 22:59, Redpanda’s Tyler Akidau says enterprises want agents as “a digital workforce” collaborating with people across every system, which is why their data access and telemetry can’t sit in a walled garden. The episode’s title, “We Built Agents, Nobody Built HR,” frames the gap: agents need the identity, permissions and oversight that people get. In episode 12, at 17:39, Logan Kilpatrick picks up Conor’s analogy of a “junior digital employee” and says the scaffolding built for text-heavy agents will need rebuilding for multimodal, agentic work. The episode was sponsored by Galileo, which Conor mentions at 18:07. Both are the guests’ views.

Editorially, the metaphor helps when it reminds a team that an agent needs a scoped job, limited permissions and a manager who checks its work, the job of human-in-the-loop review. It misleads when it implies the agent has a person’s judgment or accountability. For what agents can and can’t take on, see Can an AI Agent Run Your Workday?.