❄️
Data Flakes
Back

Getting value from AI tools requires more than access to a model. It requires understanding what the models are actually capable of, where they fall short, and how to build reliable workflows on top of them.

Anthropic has published a free learning platform — Anthropic Academy at anthropic.skilljar.com — covering everything from first-time Claude usage through to the Model Context Protocol and building production applications with the API.

Whether you are an individual contributor upskilling on your own, or a team lead trying to bring your organisation up to a consistent baseline, there is a structured learning path here worth knowing about.

What is Anthropic Academy?

Anthropic Academy is Anthropic’s official learning platform built on Skilljar. It now hosts over 20 individual courses, plus a dedicated learning path for K-12 educators, spanning beginner through to developer-level content, all free to access.

The courses are grouped loosely into four audiences:

  • Individual learners — people new to Claude or wanting to use it more effectively in their day-to-day work
  • Developers and builders — engineers working with Claude Code, the Claude API, Model Context Protocol, or agent frameworks
  • Educators, students, and nonprofits — domain-specific AI fluency programmes, now including a dedicated K-12 educator path
  • Small businesses — a fluency track aimed at non-technical teams adopting Claude
graph TD A[Anthropic Academy] --> B[AI Fluency Tracks] A --> C[Developer Tracks] A --> D[Specialist Tracks] A --> E[Claude Tools] B --> B1[AI Fluency: Framework & Foundations] B --> B2[AI Capabilities and Limitations] C --> C1[Building with the Claude API] C --> C2[Introduction to Model Context Protocol] C --> C3[Model Context Protocol: Advanced Topics] C --> C4[Introduction to Agent Skills] C --> C5[Introduction to Subagents] D --> D1[AI Fluency for Educators] D --> D2[AI Fluency for Students] D --> D3[AI Fluency for Nonprofits] D --> D4[AI Fluency for Small Businesses] D --> D5[AI Fluency for Builders] D --> D6[AI Fluency for pK-12 Educators<br/>learning path] D --> D7[Teaching AI Fluency] D --> D8[Claude with Amazon Bedrock] D --> D9[Claude on Google Cloud] E --> E1[Claude 101] E --> E2[Claude Code 101] E --> E3[Claude Platform 101] E --> E4[Introduction to Claude Cowork] E --> E5[Claude Code in Action]

The Courses

Claude 101

The natural starting point for anyone new to Claude. This course covers the fundamentals of prompting, what Claude is designed to do, and how to have productive conversations with the model.

It is well-suited for non-technical users — business analysts, operations staff, PMs — who need a practical on-ramp without getting into API mechanics.

Claude Code 101

A newer addition teaching how to use Claude Code effectively in your daily development workflow. This is the on-ramp course; Claude Code in Action (below) is the hands-on follow-up for more advanced usage.

Claude Platform 101

Teaches developers to build on the Claude Platform from the ground up — pitched at anyone from “made a few API calls” to “only used Claude through a chat window.” Useful groundwork before the more specific developer tracks.

Introduction to Claude Cowork

Covers Claude Cowork, a newer product for working alongside Claude on your real files and projects. The course is hands-on, covering the Cowork task loop, plugins and skills, file and research workflows, and how to steer multi-step work responsibly.

AI Fluency: Framework & Foundations

One of the more broadly applicable courses on the platform. This is designed as an organisational baseline — the kind of shared mental model that helps teams collaborate effectively when AI is part of the workflow.

It covers:

  • What large language models are and how they work conceptually
  • How to evaluate AI outputs critically
  • Where AI adds genuine value versus where it introduces risk
  • Frameworks for responsible use in a professional context

This is the course I would recommend running through an entire team before deploying any AI tooling. It sets expectations accurately, which is half the battle.

AI Capabilities and Limitations

A companion to the Fluency course, focused specifically on where current models excel and where they reliably fail. Understanding model limitations is underrated — teams that skip this step tend to over-trust outputs in high-stakes situations.

Building with the Claude API

For engineers building applications on top of Claude. The course walks through:

  • API authentication and basic requests
  • Structuring prompts programmatically
  • Managing context windows and conversation history
  • Handling responses and errors

This is a practical foundations course rather than a deep dive — it gets you from zero to a working integration without assuming prior LLM development experience.

Introduction to Model Context Protocol

Model Context Protocol (MCP) is Anthropic’s open standard for connecting AI models to external tools, data sources, and APIs. It is increasingly the architecture of choice for production AI workflows.

This introductory course covers:

  • What MCP is and why it exists
  • The server/client model
  • How tools and resources are defined
  • Basic MCP server setup

If you are building anything beyond a simple chat interface — agents that call APIs, read files, or interact with databases — MCP is worth understanding properly. This course provides that grounding.

Model Context Protocol: Advanced Topics

The follow-on to the introduction. Covers more complex MCP scenarios including multi-server setups, tool composition, and production deployment considerations.

Introduction to Agent Skills

Claude Code and similar agentic frameworks rely on skills — modular, reusable capability definitions that extend what an agent can do. This course introduces the concept and covers how to author and structure skills effectively.

Introduction to Subagents

Covers the subagent pattern: how to decompose complex tasks across multiple specialised agents that coordinate toward a shared goal. This is the architecture behind most production multi-agent systems.

Claude Code in Action

A hands-on course focused specifically on Claude Code — Anthropic’s agentic coding assistant. Covers real workflows for software development, code review, refactoring, and multi-file editing using Claude Code in the terminal and IDE integrations.

For any engineering team evaluating or already using Claude Code, this is the most directly applicable course.

Cloud Platform Courses

Two courses cover running Claude through cloud provider integrations:

  • Claude with Amazon Bedrock — Using Claude models via AWS Bedrock, including IAM setup, API access, and Bedrock-specific features
  • Claude on Google Cloud — Equivalent coverage for GCP, including Vertex AI endpoint configuration and authentication

These are particularly useful for teams where cloud governance requires all model access to go through an approved cloud vendor rather than directly through the Anthropic API.

Specialist Tracks

Several audience-specific AI Fluency tracks cover the same foundational material adapted for context:

  • AI Fluency for Educators — Integrating AI into teaching practice
  • AI Fluency for Students — Using AI as a learning aid responsibly
  • AI Fluency for Nonprofits — Applying AI to mission-driven work with resource constraints
  • AI Fluency for Small Businesses — Adopting Claude without a dedicated technical team
  • AI Fluency for Builders — Aimed at founders and early builders rather than enterprise teams
  • AI Fluency for pK-12 Educators — a dedicated learning path (rather than a single course) covering AI fluency for primary and secondary school teaching staff

Teaching AI Fluency is aimed at people who will be running internal AI training — useful if you are building an internal enablement programme.

Suggested Learning Paths

For an Individual Getting Started

A logical progression for someone building personal AI fluency:

graph LR A[Claude 101] --> B[AI Fluency: Framework & Foundations] B --> C[AI Capabilities and Limitations] C --> D[Building with the Claude API]

Start with Claude 101 to build intuition, move to the Fluency course for a conceptual framework, then add the Capabilities course to understand the edges. If you write code, add the API course last.

For a Development Team

Engineers and technical leads building AI-powered products:

graph LR A[Claude Platform 101] --> B[Building with the Claude API] B --> C[Introduction to MCP] C --> D[MCP: Advanced Topics] D --> E[Introduction to Agent Skills] E --> F[Introduction to Subagents] F --> G[Claude Code 101] G --> H[Claude Code in Action]

Start with Claude Platform 101 if the team hasn’t built on the platform before, then work through the API course to establish shared vocabulary, then move through the MCP and agent architecture courses in order. Claude Code 101 and Claude Code in Action form a practical capstone for the engineering track.

For an Organisation Rolling Out AI

Teams introducing AI tooling across non-technical staff:

graph TD A[Start] --> B[AI Fluency: Framework & Foundations] B --> C{Role?} C -->|Technical| D[Building with the Claude API] C -->|Non-technical| E[Claude 101] C -->|Trainer| F[Teaching AI Fluency] D --> G[MCP + Agent Courses] E --> H[Day-to-day Claude usage] F --> I[Run internal AI training]

Run the Fluency foundations course across all staff first. Then branch by role — technical staff into the API and agent tracks, non-technical staff into Claude 101, and anyone running internal enablement into the Teaching AI Fluency course.

Why This Matters for Data Teams

Data engineers, analysts, and platform teams are increasingly encountering AI in their workflows — whether through AI-assisted SQL generation, automated data quality checks, or Claude being embedded in internal tools.

The MCP and agent courses are particularly relevant here. MCP is the mechanism by which Claude can connect to a Snowflake instance, execute queries, read dbt models, or interact with a data catalogue. Understanding how that plumbing works makes you a more effective collaborator with the AI systems your organisation is building.

The AI Capabilities and Limitations course is also worth highlighting for data professionals specifically. Data work involves nuanced judgement — whether a metric is being calculated correctly, whether a data model makes sense, whether a result looks right. Knowing where Claude genuinely helps versus where hallucination risk is elevated informs how much human review you need in any given workflow.

Getting Started

The platform is free and requires only an account to access. There are no prerequisites listed for any course, and the introductory material is approachable for non-technical learners.

Visit anthropic.skilljar.com to browse the full catalogue and enrol.

From experience, I would recommend considering the following when using this as a team resource:

  1. Start with the Fluency course before any role-specific tracks — a shared conceptual baseline prevents the misaligned expectations that cause most AI rollout friction
  2. Pair the Capabilities and Limitations course with your first production use case — it is most useful when you have concrete workflows to evaluate against
  3. Use the MCP courses to evaluate your architecture, not just learn the protocol — the framing they provide is useful even if you are not building MCP servers yourself

Conclusion

Anthropic Academy is a well-structured free resource for both individual learning and organisational AI enablement. The catalogue has grown past 20 courses plus a dedicated K-12 learning path, covering a sensible range from introductory fluency through to production-grade agent architecture — and the cloud platform courses make it directly applicable for teams running Claude through AWS or GCP.

If your team is using Claude and has not yet gone through any structured training, the Fluency and Limitations courses alone would meaningfully improve the quality of how that tool gets used day-to-day.

Key Takeaways:

  • Anthropic Academy at anthropic.skilljar.com hosts over 20 free courses plus a dedicated K-12 educator learning path
  • Courses span beginner AI fluency through to MCP, agent skills, Claude Code, and the newer Claude Cowork product
  • Cloud platform courses cover Bedrock and Google Cloud (Vertex AI) integration specifically
  • Suggested paths: individual upskilling, development teams, and organisation-wide rollout
  • MCP and agent architecture courses are particularly relevant for data engineering teams

Further Reading

Disclaimer

The information provided on this website is for general informational purposes only. While we strive to keep the information up to date and correct, there may be instances where information is outdated or links are no longer valid. We make no representations or warranties of any kind, express or implied, about the completeness, accuracy, reliability, suitability, or availability with respect to the website or the information, products, services, or related graphics contained on the website for any purpose. Any reliance you place on such information is therefore strictly at your own risk.