Turn Claude Code into a fleet of AI agents that do real procurement and supply-chain work, no coding background needed
For procurement professionals, supply-chain analysts, and students who want to build working AI agents from scratch and walk into their next interview able to show them running.
16 chapters, 18 lessons
14-day refund on the yearly plan. Real pricing on the plans page, no surprises.
18
lessons, yours to run
Easy to pick up, built to get you the next job
No coding background needed, and you write code with Claude as you go.
Every section below turns into something a hiring manager recognises, not just notes.
Go from zero to a running agent in your first session
- You will install Claude Code, set up your workspace, and run a working agent on day one, following plain step-by-step instructions with no prior coding required.
- You will write CLAUDE.md instruction files that give an agent a defined role, hard rules, and a no list, exactly the kind of controlled automation a hiring manager trusts.
- You will understand what makes an agent different from a chatbot, a distinction that comes up in every AI-related job description right now.
Design agents by making real decisions, not watching someone else build
- You will work through fifteen industry cases, each one starting with a short business story and a guided design interview where every architecture choice is yours to make.
- You will practice the design thinking that separates an AI operator from someone who just prompts, a skill set employers are actively hiring for.
- You will build each case in Claude Code with a companion starter repo, so you finish with working artefacts, not just notes.
Connect agents to real tools and data sources
- You will build agents that use MCP tools to reach live systems, turning a standalone script into something that can query, act, and report.
- You will wire up subagents that hand work between each other, so a single prompt can trigger a multi-step workflow across different data sources.
- These are the capabilities listed in AI engineer and operations automation job specs, and you will have them running as portfolio proof.
Write real agent code with the Claude Agent SDK
- You will progress from instruction files and hooks into actual agent code using the Claude Agent SDK, guided line by line so the jump never feels steep.
- You will build multi-agent systems where specialised agents collaborate on a task, the architecture behind the most in-demand automation roles today.
- Every piece of code you write is grounded in one of the fifteen cases, so you finish the course able to demo a real scenario, not a toy example.
Finish this course and you can do all of this, no prior background required:
- You will be able to design and build a working AI agent from a business brief, the core skill in every AI operations and automation role being hired for right now.
- You will be able to write structured instruction files and no lists that make an agent predictable and auditable, exactly what a team lead needs before they trust automation in a live workflow.
- You will be able to connect agents to real tools via MCP so they can query systems, take actions, and return results, moving you from prompt writer to AI operator.
- You will be able to architect multi-agent systems where specialised subagents collaborate on a task, a capability that appears in senior AI and automation job descriptions and that almost no candidate
- You will be able to write real agent code with the Claude Agent SDK, giving you the technical credibility to work alongside engineering teams and contribute to production builds.
- You will be able to walk into an interview with a portfolio of fifteen running agents built on real business cases, turning every question about AI experience into a live demo.
18 lessons, 16 chapters
Built by practitioners, not influencers
Every agent, skill file, and dataset is drawn from 25+ years and 50+ Oracle ERP implementations across pharma, manufacturing, semiconductor, and distribution.
Kept current with the stack
Claude Code, skills, and MCP move fast. Your licence includes every system update for as long as you're subscribed. Your system gets better, not obsolete.
You can see that AI is changing how operations teams work, but every course you find assumes you already know Python or asks you to follow along while someone else types. You end up with a certificate and nothing running. Meanwhile, job postings for AI-enabled procurement and supply-chain roles are piling up, and the candidates getting interviews are the ones who can show an agent doing something real. If you cannot point to a working build, you are invisible to those hiring managers. Watching demos and reading threads is not going to close that gap.
What you'll be able to do, module by module
Every module leaves you able to build and run something real. Here is the syllabus, in the order you work through it.
Curriculum
16 chapters · 18 lessons
Part 1: Get Ready
You can install Claude Code, connect it to your account, create a scoped workspace, and ship a working instruction-file agent in a single session, with no prior coding experience and a clear troubleshooting path if anything goes wrong.
Before You Build4 items
- L1: Set Up Claude Code and Build Your First Agent in 10 Minutes🔒
- L2: The Seven Building Blocks You Can Actually Touch🔒
- L3: How This Course Works + the Repo Tour🔒
- Quiz: Orientation (Setup, the Seven Blocks, and the Course Method)Quiz
Part 2: First Builds with Claude Code
You can take a real business case, work through a guided design interview to make the architecture decisions yourself, then build and run the agent in Claude Code using a companion starter repo, finishing with a working artefact you own.
DC-01: Uber: The Router You Build in an Afternoon1 item
- DC-01: The Router You Build in an Afternoon🔒
DC-02: Swiggy: Two Seconds to Say Something True1 item
- DC-02: Two Seconds to Say Something True🔒
DC-03: Amazon: The Agent With the Keys to the Money1 item
- DC-03: The Agent With the Keys to the Money🔒
DC-04: Airbnb: The Judge You Build Behind a Wall1 item
- DC-04: The Judge You Build Behind a Wall🔒
DC-05: IndiGo: The Storm Desk You Build in an Evening2 items
- DC-05: The Storm Desk You Build in an Evening🔒
- Quiz: Tier 1 (CLAUDE.md, Subagents, Slash Commands, Guardrails as Instructions)Quiz
Part 3: Agents with Real Tools
You can connect an agent to live systems using MCP tools, structure multi-step workflows with subagents that hand work between each other, and use hooks to control agent behaviour, producing automations that go beyond a single prompt and answer.
DC-06: Razorpay: The Defense Packet You Build Against the Clock1 item
- DC-06: The Defense Packet You Build Against the Clock🔒
DC-07: Duolingo: The Tutor That Refuses to Help1 item
- DC-07: The Tutor That Refuses to Help🔒
DC-08: LinkedIn: The Screener That Shows Its Work1 item
- DC-08: The Screener That Shows Its Work🔒
DC-09: Datadog: The Agent That May Look but Not Touch1 item
- DC-09: The Agent That May Look but Not Touch🔒
DC-10: Practo: The Triage Router That Never Diagnoses2 items
- DC-10: The Triage Router That Never Diagnoses🔒
- Quiz: Tier 2 (Hooks, MCP, Multi Agent Delegation, Audit Trails)Quiz
Part 4: Real Agent Code with the Claude Agent SDK
You can write actual agent code using the Claude Agent SDK, build multi-agent systems where specialised agents collaborate on a shared task, and produce a complete, demonstrable build that shows a hiring manager you can operate at the level the role requires.
DC-11: Walmart: The Negotiator That Knows When to Stop1 item
- DC-11: The Negotiator That Knows When to Stop🔒
DC-12: Support at Scale: The Router That Watches the Meter1 item
- DC-12: The Router That Watches the Meter🔒
DC-13: Perplexity: The Researcher That Shows Its Sources1 item
- DC-13: The Researcher That Shows Its Sources🔒
DC-14: HDFC: The Vault You Build Around the Model1 item
- DC-14: The Vault You Build Around the Model🔒
DC-15: Ola: The Night Watcher You Build for Real2 items
- DC-15: The Night Watcher You Build for Real🔒
- Quiz: Tier 3 (the Claude Agent SDK: Tools in Code, Guardrails in Code, Evals)Quiz
Candidates who can design, build, and demo a working AI agent are rare enough that hiring managers remember them. Completing this course puts you in the small group who can answer every AI question in an interview by opening a terminal and showing the thing running.
From subscribing to running your first artifact
Subscribe and get instant access
Subscribe to the academy, log in, and every lesson, starter repo, and design interview is available immediately. No software to configure before you begin, just open the first lesson.
Open Part 1 and follow the plain steps
Lesson one walks you through installing Claude Code, connecting your account, and setting up your first workspace. Every instruction tells you exactly what to type and what you should see when it works. If it does not match, there is a troubleshooting section. No coding background assumed, no step skipped.
Build your first agent and keep building through all fifteen cases
By the end of lesson one you have a working agent running from two text files you created yourself. From there you move through fifteen real business cases, each one a story, a design interview, and a guided build in Claude Code with a companion starter repo. You finish with a portfolio of running agents you can show to any hiring manager.
Building a supplier onboarding triage agent, step by step
Read the business story and identify the problem
The case presents a procurement team drowning in supplier onboarding requests, each one requiring the same sequence of checks before a buyer even looks at it. You read the story and identify the repetitive decision points that an agent could own.
Work through the design interview
The guided design interview asks you to decide: what is the agent's goal, what information does it need to act, what tools does it require, and what must it never do without a human in the loop. You make every call. This is the thinking a hiring manager wants to see you do.
Write the instruction file and define the no list
You create a CLAUDE.md that gives the agent its role, its decision rules for classifying a request as complete, incomplete, or escalate, and a hard no list that prevents it from approving a supplier without a compliance check. The no list is what makes the agent safe to deploy.
Build and run the agent in Claude Code with the starter repo
You open the companion starter repo, follow the step-by-step build instructions in Claude Code, and run the agent against a sample set of onboarding requests. It triages each one, flags missing documents, and routes escalations, all without you touching every case manually.
Review the output and iterate
You check the agent's decisions against the expected outcomes, adjust the instruction file where the logic misfired, and rerun. You finish the case with a working triage agent and a clear record of the design decisions behind it, exactly what you would walk a hiring manager through.
The toolkit you'll build and run in this program
You leave with working assets, not notes.
Real things you can demo in an interview.
Fifteen working agents built on real business cases, each one demonstrable in a live interview or portfolio review.
A companion starter repo for every case, giving you a clean foundation you can extend or repurpose for your own organisation.
CLAUDE.md instruction file templates covering role definition, operating rules, and no lists, ready to adapt to any workflow.
Multi-agent system builds using the Claude Agent SDK, showing end-to-end architecture from prompt to action.
MCP tool integrations that connect agents to live systems, portfolio proof that you can automate beyond a single chat session.
You'll get the most from this if
Procurement and supply-chain professionals who want to automate repetitive operational work and make AI a visible part of their role.
Analysts and coordinators who handle high-volume tasks like supplier checks, document triage, or workflow routing and want to hand the heavy lifting to an agent.
Students studying procurement, operations, or business who want a practical AI skill that makes their CV stand out before they have years of work experience.
Career changers who understand operations workflows and want to add AI agent building to their toolkit without needing a software engineering background.
Yearly plan: 14-day keep-the-system guarantee
14-day refund on the yearly plan. Finish the first two modules, and if you haven't run your first live pipeline, take a full refund.
Straight answers before you subscribe
Do I need any coding or technical background to start?
No. The course is built on the premise that you have none. Lesson one tells you exactly what to type, names what you should see after every step, and includes a troubleshooting section for when it does not match. Every subsequent lesson follows the same pattern. You learn by building, not by reading theory.
Will this work with my current computer setup?
Claude Code runs on macOS and Windows. You need an internet connection and a Claude paid plan such as Pro or Max, because the free plan does not include Claude Code. The course covers the install from scratch, so if your machine meets those three requirements you are ready to start.
What if I get stuck or fall behind?
The academy runs daily live sessions where you can bring your build and get it looked at in real time. There is also mentor review, so you are not debugging alone. The course is designed for people learning alongside a job, not for people with unlimited hours, and the support structure reflects that.
How does the course stay current as Claude Code changes?
The course is updated as the tools evolve. When you subscribe you get access to updated lessons, not a frozen snapshot. If a step changes because Claude Code released a new version, the lesson reflects it.
What is the refund policy, and can I use what I build commercially?
There is a 14-day refund period. If the course is not what you needed, request a refund within 14 days and it is honoured. The agents and artefacts you build during the course are yours. You can use them in your work, adapt them for your organisation, or show them in a portfolio or interview.