Chainbook

Build intelligence.
Train for the future.

Chainbook equips professionals, developers, and businesses with the knowledge to design, deploy, and govern Artificial Intelligence systems that actually work in the real world.

Get in touch

Our approach

Practical knowledge for
intelligent systems.

We don't teach AI as theory. Every course, guide, and framework we publish is grounded in real-world application — the same systems Chainbook builds for accountants, property managers, and financial advisors across Africa.

Who it's for

Developers, founders,
and professionals.

Whether you're a software engineer designing your first agent, a business owner evaluating AI adoption, or a professional who wants to understand the systems shaping your industry — the Academy has a path for you.

Featured framework

Framework for Agentic AI Systems Design

Companies are consistently adopting AI-driven systems through the use of AI Agents, yet most developers find themselves without a clear, scalable framework for designing and building them. This framework defines the exact steps — from scoping an agent's purpose to selecting protocols and enforcing constraints — to produce systems that are both maintainable and scalable.

Step 01

Define the Agent

Clearly scope what the agent does and identify every AI touch point and interaction. This stage governs which protocols you will need — and getting it wrong cascades into every decision that follows.

Step 02

Select Protocols

Choose from the six core agent protocols — MCP, A2A, UCP, AP2, A2UI, and AG-UI — based on how your agent interacts with data, other agents, commerce systems, and user interfaces.

Step 03

Design the Interaction Layer

Map each identified action to the protocol that best achieves it. Ask whether a task requires agent-to-agent delegation, data retrieval, a commerce flow, or a frontend interface — then engineer accordingly.

Step 04

Apply Constraints

Govern agent behaviour beyond protocols. Define safety boundaries, data privacy guardrails, and security controls — so the system acts with the same rigour you'd expect from any intelligent business tool.

Get in touch

The six core protocols

How intelligent agents communicate.

Context-oriented

Model Context Protocol

A single standard connection pattern that lets agents discover tools and access data sources — databases, files, APIs — without custom integration code for every service.

Inter-agent

Agent-to-Agent Protocol

Standardises how agents discover and communicate with one another, enabling a primary agent to delegate tasks to remote specialist agents built on different frameworks.

Domain-specific

Universal Commerce Protocol

Standardises the shopping lifecycle — catalog discovery to checkout — into modular capabilities, letting an agent interact with multiple suppliers using a single unified pattern.

Domain-specific

Agent Payment Protocol

Handles payment authorisation with typed mandates and configurable guardrails. Works alongside UCP — UCP manages commerce logistics while AP2 manages authorisation and audit trails.

Interfacing

Agent-to-UI Protocol

Allows agents to dynamically compose native UI layouts — dashboards, forms — from safe component primitives, separating structure from data without requiring custom frontend code.

Interfacing

Agent-User Interaction Protocol

Acts as middleware that translates agent events — streaming text, tool calls — into a standardised stream, eliminating client-side parsing boilerplate when rendering agent responses.

Ready to get started?

Let's build something intelligent together.

Interested in AI training for your team, or want to explore how intelligent systems can transform your business? We'd love to hear from you.

Get in touch