From Assisted to Agentic AI Development

• Mitigating future friction with rules
• Document-driven development patterns
• Building the start of a context framework
• Practice structured planning sessions
• Multi-turn interactions for complex problems
• Debugging and refactoring with an assistant
• Investment mindset: treating AI setup like IDE optimisation
• Project Continue: Complete Configuration API with comprehensive tests and docs

• Creating reusable templates and scaffolds
• Building with context assets and examples
• Establish positive feedback loops for your assistant
• Combining static and dynamic context
• Project Complete: Complete Configuration API with comprehensive tests and docs

• Implementing and HTTP MCP server from scratch
• Wrapping existing servers (your Configuration API)
• Error handling and production patterns
• Project: Build and deploy your own MCP server

• Sequential orchestration: Chaining MCP tools for complex workflows
• Parallel coordination: Running multiple tools simultaneously
• Conditional logic: Dynamic tool selection based on context
• Error handling: Fallback strategies and recovery patterns
• Observability: Monitoring multi-tool operations
• Project: Coordinate multiple MCP servers to automate a development workflow

• Agent communication patterns
• Specialisation and role-based design
• Shared context and memory management
• Conflict resolution between agents
• Project Evolve: Create a multi-agent system for development workflow automation

• Monitor and measure agent system performance
• Optimise agent decision-making and efficiency
• Build feedback loops for agent improvement
• Understand the impact of agents on development workflows
• Build feedback loops for agent improvement
• Project Complete: DORA and Flow Metrics dashboard

• Lead with empathy through transformation
• Change management principles
• Addressing adoption objections
• Building communication matrices and plans

• Understand the ‘why’: Through the course learning objectives, we address how AI engineering practices improve individual and team flow, reduce technical debt and improve quality. We also explore ways to mitigate the potential risks and costs associated with adoption
• Pedagogy and approach: Our unique teaching method focuses on building an intuition around AI tools and their application, rather than just providing step-by-step instructions
• Resilience to change: You will learn a framework for understanding and adapting to future changes in AI, ensuring your learning remains relevant and your skills resilient
• Skills relevance: Combine your software engineering experience with AI skills to stay relevant in the market and maintain job security
• Beyond code completion: This course goes far beyond basic code completion, teaching you how to learn and use it for resilience in the changes coming to software development practices
• Addresses AI-coding myths and acknowledges realities:
• Myth: Senior devs are more productive without it
• Myth: AI-assisted software development is all about speed
• Myth: AI coding tools are just spicy autocomplete that speed up typing
• Myth: To develop software with AI support means you must “vibe-code”
• Reality: You won’t always get a productive experience “straight out of the box” with AI-coding tools. There are many practices and constraints to learn
• Reality: You can risk ceding too much control to the models if you’re not careful
• Reality: Things are moving at such a rapid pace that you cannot pin down ‘best practices’ but rather must identify concepts and knowledge you can use to help apply judgment as things evolve
• Practical application: The course emphasises building a set of practices for successfully leveraging AI tools. You’ll learn to identify and apply emerging patterns for effective AI-assisted development deliberately
• Flexible learning: While there will be dedicated live lecture slots, video recordings will be available for flexible viewing at your convenience. Similarly, capstone project exercises can be adapted to your schedule

What will I learn in the From Assisted to Agentic AI Development course? This course takes you through a hands-on journey from basic AI collaboration to building fully autonomous AI systems. In practical terms, you’ll learn how to build production-ready software with AI assistance, automate development workflows using AI tools and memory techniques and design multi-agent AI systems that can work together to solve problems.

The curriculum is organised in three phases and, as you progress, you’ll gain skills in prompt engineering, tool integration (like function calling and retrieval systems) and orchestrating intelligent agents. By the end, participants even learn how to lead AI adoption in their organisations, combining the technical know-how with change management principles for introducing AI into teams.

What kind of projects will I work on? You’ll build several real-world projects including a multi-tenanted microservice with AI-assisted lifecycle (including planning and documentation), a domain-specific assistant with RAG and memory, automated dev workflows and AI reviews and research and writing agents collaborating at scale. Along the way, you’ll build real projects which will give you portfolio pieces and experience with AI in real-world scenarios.

Will these projects be portfolio-worthy? Definitely. All projects are designed to give you tangible portfolio pieces and real-world experience with AI in software development scenarios.

What is ‘AI-assisted development’? AI-assisted development refers to using AI tools, such as Claude Code, to help software developers write, review and optimise code more efficiently. These tools leverage machine learning models trained on codebases to make recommendations.

When combined with high-quality examples and relevant context, these tools can generate code, tests, infrastructure and other important artefacts, speeding up development while improving quality.

What does ‘agentic AI development’ mean? Agentic AI development is about creating AI systems (or ‘AI agents’) that act with autonomy and initiative, rather than just following a single prompt or rule. In practice, agentic AI systems can make independent decisions and perform multi-step tasks to achieve goals with appropriate oversight.

This means moving beyond using AI as a passive assistant. Developers can build orchestrated, autonomous AI agents that plan, execute and adapt to complete complex objectives independently, provided deterministic guardrails are in place to check results.

How can AI tools benefit software developers in coding? AI tools can dramatically boost a developer's productivity and quality by reducing toil, accelerating the implementation of known patterns and providing intelligent search and guidance.

For instance, AI coding assistants can accelerate prototyping, generate code based on appropriate patterns, recommend and apply refactorings to reduce technical debt, help with code analysis, accelerate code migration, generate high-quality documentation, coach junior developers and broaden individuals' skills, allowing for teams to be more cross-functional and autonomous.

When used effectively, the result is faster and higher-quality software, allowing more time for humans to focus on creative and complex aspects of software design and delivery.

How will this course impact my daily development work? You’ll dramatically boost your efficiency by learning to automate tedious coding tasks and leverage AI for intelligent guidance. The skills you learn will help you focus more on creative and complex aspects of software design.

Why should tech leads and engineering managers invest in AI development training for their teams? Adopting AI in the development process is quickly becoming a priority for organisations, with most tech leaders planning on expanding the use of AI in their software teams in the near term.

By investing in AI engineering training, tech leads and engineering managers ensure their teams stay ahead of the curve and can fully leverage AI tools for productivity and innovation. A structured programme equips engineers to build production-ready AI systems and even lead AI adoption and change management inside teams.

Training developers in AI improves day-to-day efficiency and empowers your engineers to drive the company’s AI strategy effectively.

How can companies support AI adoption in their software development teams? Companies can support AI adoption by upskilling their developers and fostering a culture of psychological safety where engineers can experiment and learn new techniques, combined with targeted training courses.

Cohort-based training helps engineers learn how to use AI tools and build agentic systems effectively in small groups, rather than expecting them to figure it out independently. It’s also important to designate AI champions or leaders who can drive developer enablement, experts who advocate for AI use, share best practices and mentor others.

From a change management perspective, leadership should introduce AI with empathy, addressing developers’ concerns (like fear of job impact or trust in AI outputs) and clearly communicating the benefits and goals of using AI.

By creating an environment that encourages experimentation with AI, provides the necessary tools and addresses challenges, leaders can create the conditions for AI adoption and accelerate benefits.