Hands-on workshops

Context systems often fail in ways that better prompting alone will not fix. In this hands-on workshop with the author of O'Reilly's The Context Engineering Handbook, you’ll diagnose a range of common—and less common—context failures, understand why they occur and implement evidence-based fixes with DSPy.

You’ll build a context-engineering harness that can evaluate performance, expose failures and support iterative improvement. Using compound-AI techniques, you’ll create compaction functions tailored to particular use cases, add security mechanisms for coding agents and build tool-selection functions that reduce context bloat. The emphasis is on developing useful mental models as well as working code.

You’ll leave with a starter repository and a working harness you can continue adapting to your own systems.

Who it’s for

Software and AI engineers who want to build more reliable context systems. Python fluency is helpful, but participants who work in another language and are comfortable using a coding agent should be able to follow. No previous DSPy experience is required.

What you’ll need

Bring a laptop on which you can install and run a local Python development environment. You’ll need GitHub access, uv installed, a coding agent and a model API key from an inference provider. Enterprise-managed laptops that prevent you from creating a Python environment are unlikely to work.

Can an AI security reviewer find vulnerabilities it has never seen—or is it simply repeating familiar benchmarks? In this three-hour hands-on workshop, you’ll build a security-review agent and the evaluations needed to tell whether it actually works.

Starting with a supplied TypeScript agent, you’ll examine how its context, tools, prompts and structured outputs shape the result. You’ll run a baseline evaluation, inspect the complete execution trace and investigate benchmark contamination: whether apparently strong performance comes from genuine reasoning or familiarity with public vulnerability data.

You’ll then create replayable test fixtures derived from real CVEs across several vulnerability classes, compare model and harness configurations using recall, runtime and cost, diagnose failed or incomplete results and improve the system against evidence.

You’ll leave with a working security-review agent, CVE-backed evaluation fixtures, a method for separating model failures from harness failures, example vulnerability cases and workshop notes you can adapt to your own language and security work.

Who it’s for

Intermediate AI engineers, application-security engineers and software engineers. You should be comfortable reading TypeScript and working in a terminal. Familiarity with application-security concepts or an agent framework such as the Vercel AI SDK will help, but is not required.

What you’ll need

Bring a laptop with Docker configured, GitHub access and a coding-agent subscription such as Claude Code, ChatGPT or Cursor. Setup and repository instructions will be provided before the workshop.

Give an agent a goal in plain language. In this hands-on workshop, you’ll use BAML to build an agent that creates the tools it needs on demand, critiques its own failures, rewrites those tools and improves across successive iterations.

You’ll then take the same loop toward production: mining real execution traces for failures and feeding them back into continuous tool improvement. By the end, you’ll have built a working self-improving agent, implemented a feedback loop based on failed runs, and connected that loop to a post-deployment improvement workflow.

You’ll leave with a working repository containing all workshop code under an open-source licence. BAML supports OpenAI, Anthropic, open-source models, Vertex AI and other model providers, so the techniques are not tied to one model.

AI-assisted development is reaching every part of software engineering—including embedded development on microcontrollers such as the ESP32. In this introductory, hands-on workshop, participants will connect an AI coding harness to M5Stack ESP32 modules and build an AI-themed project from scratch.

You’ll progress from setup and a first upload to creating a personal voice-powered recorder and assistant.

By the end, you’ll have a working project, a physical module to take home, and the foundations to continue exploring AI-assisted IoT development yourself.

No previous Arduino, PlatformIO or ESP32 experience is required.

AI agents can inspect and change cloud infrastructure—but how do you let them act without giving up control?

In this hands-on workshop, you’ll first build a read-only inventory and audit agent that queries live resources in an isolated cloud environment and reports what is deployed. You’ll then choose a specialisation—FinOps, site reliability engineering, cloud security posture management, or entitlements and access review—and extend the agent to identify issues and recommend a response.

Finally, you’ll move safely from observation to action: the agent proposes a change, a human approves it, and the agent executes and verifies the result.

You’ll leave with a working cloud-operations agent and access to a public MIT-licensed repository containing every workshop track. No personal cloud or model account is required; isolated environments, credentials and model access are supplied.

Claude Code, Codex and other AI coding agents are transforming how developers work. But their cost, privacy implications and dependence on proprietary models are leading many teams to explore more open alternatives.

In this three-hour hands-on workshop, you’ll recreate an agent-assisted development workflow using the open-source OpenCode harness and ResetData-hosted open-weight models running entirely on Australian AI infrastructure.

You’ll configure OpenCode and connect it to an open-weight model; turn a development goal into clear requirements and a practical “done when” test; and use slash commands, skills and hooks to plan and build software with a human in the loop. You’ll then add automated tests to check the result.

We’ll also look at how ResetData uses Claude and OpenCode side by side: what can move successfully to an open stack, where proprietary models may still be useful, and how open- and closed-weight models can form part of the same development workflow.

You’ll leave with a working OpenCode setup and a repeatable plan–build–test workflow that you can continue using with open-weight models.

Who it’s for

Software developers who already use—or are actively evaluating—AI coding agents. No previous OpenCode or open-weight-model experience is required.

What you’ll need

Bring a suitable development laptop. You’ll use OpenCode, Visual Studio Code and a ResetData account created before the workshop. ResetData will provide the required model access and token voucher at no additional cost. Exact installation steps and supported operating systems will be confirmed before the event.