Skip to content

Training by Amzu · Online course

Build with AI agents.
Ship like an enterprise.

A hands-on training in secure, production-ready, enterprise-grade full-stack development with AI coding agents — covering Claude Code and OpenAI Codex from first prompt to paying customers.

Claude Code & OpenAI CodexOnline — access any timeWeb-first, mobile includedSecurity & production focusTwo capstone projects

What you take away

Not another demo.
A working practice.

Most AI coding material stops at the prototype. This training is about everything after: the engineering, security, and operational discipline that turns agent-written code into software you can put in front of customers.

01 / BUILD

Ship complete products with AI agents

Take an idea to a working, deployed full-stack application using Claude Code and OpenAI Codex as your engineering team.

02 / SECURE

Engineer to an enterprise standard

Apply the security, testing, and review disciplines that separate a demo from software a business can depend on.

03 / OPERATE

Run software in production

Deploy, monitor, fix, upgrade, and monetise what you build — the whole lifecycle, not just the first commit.

The curriculum

Eight modules.
The full lifecycle.

From your first structured prompt to monitoring, upgrades, and revenue. Each module pairs concepts with hands-on work in Claude Code and OpenAI Codex.

01

AI coding agents, properly introduced

How Claude Code and OpenAI Codex actually work — context, planning, tool use, and the workflows that make agents reliable rather than lucky.

02

Choosing your full stack

Selecting languages, frameworks, and data layers that agents work well with — and that your future team can maintain.

03

Engineering best practices with agents

Specifications, version control, testing strategy, and the habits that keep AI-generated codebases coherent as they grow.

04

Security by design

Authentication, secrets, dependency hygiene, and OWASP-aligned practices — built in from the first prompt, not bolted on.

05

Deployment environments

Shipping to the cloud and to platforms like Vercel and Railway — choosing the right home for each project, with CI/CD that agents can drive.

06

Production, bug fixes & upgrades

Continuous monitoring, incident response, dependency upgrades, and using agents to maintain software — the part most courses skip.

07

Monetising what you build

Payments, subscriptions, and pricing — turning a deployed application into a product people pay for.

08

Capstone: two projects, two philosophies

A partially controlled build where you review every change, and a fully AI-led build where guardrails replace line-by-line review.

The complete session-by-session syllabus, project briefs, and tooling list are shared with the waiting list before the course opens.

What you'll build

Web-first.
Mobile-aware.

Primary focus

Web applications

The heart of the course. Full-stack web products with real authentication, data, payments, and deployment — the kind of application most businesses run on.

  • Modern frontend and API design with agent-driven workflows
  • Databases, background jobs, and third-party integrations
  • Deployed to cloud providers and platforms like Vercel and Railway
Also covered

Mobile applications

A focused look at building and shipping mobile apps with AI agents — where the workflow differs, and what changes for app-store delivery.

  • When to choose native versus cross-platform
  • Agent-assisted mobile development in practice

Two ways to work

You in the loop.
Or out of it.

The industry is split between AI-assisted engineering and fully autonomous delivery. One programme, two capstone projects — you'll practise both ways of working and learn when each one is the right call.

Project 01 / Partially controlled

The guided build

Agents propose; you decide. Every change passes through your review before it lands.

  • Code review as a core discipline — what to look for in agent-written code
  • Structured plans, small diffs, and honest test suites
  • The workflow for teams that answer to auditors and customers
Project 02 / Fully AI-led

The autonomous build

No line-by-line review. Agents plan, build, test, and ship — and your job becomes designing the guardrails they operate within.

  • Specifications, automated checks, and CI gates instead of manual review
  • Monitoring and rollback as your safety net
  • Where this works brilliantly — and where it still shouldn't be trusted
Venkatesh Balakumar, course instructor

Your instructor

Venkatesh Balakumar

CTO & Founding Director, Amzu · AI & Cloud Engineering

Venkatesh is the Chief Technology Officer and a Founding Director of Amzu, where he has architected the AI and cloud platforms behind the company's products — turning ambitious ideas into resilient, production-ready systems.

An AI, cybersecurity, and cloud strategy and development consultant for large financial services clients, he pairs deep engineering craft with a strategic view of how technology creates value. An alumnus of Saïd Business School at the University of Oxford, he works hands-on with AI coding agents every day — the same workflows this course teaches are the ones Amzu ships with.

  • AI, Cybersecurity & Cloud
  • Financial Services Consulting
  • Engineering Leadership
  • Saïd Business School, Oxford
  • Startup & Enterprise Advisor
  • Ships with AI agents daily
Meet the Amzu team

Taught from practice

We teach what
we ship.

This isn't theory assembled from other people's blog posts. The workflows in this course are the ones behind Amzu's own enterprise platforms, internal agent systems, and a portfolio of published consumer apps.

ENTERPRISE PLATFORM

Meivo

Our enterprise vector-memory platform for RAG — deployed across GCP, AWS, Azure, Kubernetes, and air-gapped environments, with tamper-evident audit and layered guardrails.

ENTERPRISE PLATFORM

PeopleHQ.ai

Context-aware AI over organisational knowledge, with enterprise-grade security, auditability, and traceability built into every interaction.

INTERNAL PLATFORM

Amzu Sales & Marketing

AI agents with defined roles coordinating prospecting, outreach, and follow-up across Apollo, Resend, and Mailgun — running our own pipeline daily.

APP STORE · CONSUMER APPS

Apps by Amzu

ZenWalk, HerFit, GrowHut's Verna, Dictaphone, and more — published on the App Store and operated in production with the same agent-driven practice.

The waiting list

Learn any time.
Be first to start.

This is a fully online, self-paced course — once it opens, you can start whenever suits you and learn at your own rhythm. Join the waiting list and you'll receive the full syllabus, launch date, and pricing before anyone else — with early access when the doors open.

  • 01Full session-by-session syllabus, sent to the list first
  • 02Early access and launch pricing when the course opens
  • 03No spam — course announcements only, unsubscribe any time
Enterprise training

We also arrange specific courses for enterprises — tailored to your team, your stack, and your schedule. Tick the enterprise option in the form and we'll be in touch.

Personal or company email — both are welcome.

This form is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.