agentnetwork.ai · April 2026

The Agent
Mapper

Six stages from chatbot to synthetic employee. Most organisations are further back than they think. Find where you are. Understand what the next stage actually requires.

01 The Prompter 02 The Augmented Worker 03 The Workflow Builder 04 The Agent Operator 05 The Orchestrator 06 The Synthetic Employer

The Six Stages

From chatbot to synthetic employee

Select any stage to explore what it looks like, which agent components are present, and what the real-world proof points are. Quick orientation: if your AI only works when you are in the room, you are at stage one or two regardless of how many tools your organisation has deployed.

The Anatomy of an Agent

A chatbot has one component. An agent has six.

Most of what organisations call an agent is a model with a chat interface. A genuine agent requires all six components below. Most deployments have one. The rest is marketing.

The pi.dev principle: A proper agent is a model plus a harness plus a heartbeat plus memory plus tools plus a soul. This framework — from the engine powering the OpenClaw ecosystem — is the clearest technical definition of a genuine agent. Most chat deployments have the model. They have nothing else.

The proof that changes everything

Block rebuilt how it works. The numbers from the CFO.

In the past eighteen months, Block used an open-source agent framework called Goose to rebuild how its entire organisation works. The results, reported directly by the CFO to Fortune, are the clearest real-world proof of what Stage 5 to 6 operation looks like.

40% developer productivity increase per engineer since Sept 2025
$2M projected gross profit per employee in 2026, up from $500K in 2019
18mo the time it took to rebuild how the entire organisation works

Block CFO Amrita Ahuja · Fortune · March 2026

Block built Goose internally, deployed it across every function, and gave it away to the world under an open-source licence. One risk underwriting model that previously took an entire quarter to build was completed in a fraction of the time. Non-technical teams now run SQL queries and close support tickets without waiting for an engineer. The scale of that transformation meant the organisation needed to reshape itself around the new capability — and in early 2026 it did, reporting its best quarter in history the same month.

Block's CTO, Dhanji Prasanna: the competitive advantage lies in the agent harness built around the models, not in the models themselves. Every organisation is focused on choosing the right model. That is the wrong question. The model is the least defensible part of the stack. Block did not win because it picked the best model. It won because it understood which layer the real advantage lives in.

The Diagnostic

Where are you actually?

Answer five questions about how you and your organisation use AI today. The diagnostic places you on the six-stage map and shows you specifically what the next stage requires.

Someone shared their badge with you

What stage are you? Take the five-question diagnostic and get your own badge.

0 of 5

Your current stage

Read the full article ↗

Your badge