Jurre Brandsen
AI Champion & Software Engineer at Info Support
AI Champion & Software Engineer bij Info Support
Utrecht, The Netherlands
Actions
Jurre is a Software Engineer at Info Support with a focus on AI augmented software engineering. With a solid foundation in machine learning, he explores how AI can practically assist developers, from improving workflows to supporting decision-making throughout the software lifecycle. Jurre enjoys exploring practical ways to make AI a natural part of everyday development, aiming for tools and techniques that feel intuitive and useful.
Jurre is Software Engineer bij Info Support met een focus op AI-geaugmenteerde software-engineering. Met een sterke basis in machine learning onderzoekt hij hoe AI-ontwikkelaars op een praktische manier kan ondersteunen, van het verbeteren van workflows tot het ondersteunen van besluitvorming tijdens de hele softwarelevenscyclus. Jurre vindt het interessant om te ontdekken hoe AI op een natuurlijke manier onderdeel kan worden van dagelijkse ontwikkelpraktijken, met tools en technieken die intuïtief en bruikbaar.
Area of Expertise
Topics
Part 2/2: Your Agent Harness Never Learns From Last Week. Let's Fix That.
Every week, the same kind of friction shows up in your agent sessions. A misunderstanding that keeps coming back. A workaround you’ve explained before. That familiar moment of: “Right, I need to tell it this again.”
In this hands-on workshop, you’ll work with an existing agent harness and build a small continuous-improvement loop around it. The routine looks across recent sessions, finds recurring patterns, and helps turn them into concrete improvements: a sharper instruction, a new rule, a skill, a guardrail, or whatever the situation calls for.
You’ll also document the change as an HDR: Harness Decision Record. What pattern triggered it? What evidence did you see? Why is this the right fix? And how will you validate whether it actually worked?
By the end, you’ll have a working improvement routine, your first HDR, and a harness that learns from repeated friction instead of leaving you to catch the same problems manually every time.
Build Your Self-Improving Agent Harness From the Ground Up
Most agentic workflows don’t fail because the model isn’t capable enough, but because the environment around it is incomplete. Important context lives in someone’s head, instructions are scattered, and the agent keeps making the same mistakes.
In this full-day workshop, you’ll build an agent harness around a project we provide. You’ll work with instructions, skills, agents, hooks, project knowledge, validation, and persistent context, and learn how these pieces work together to make an agent more reliable.
Then we’ll go one step further: making the harness you built in the morning continuously improve over time. You’ll use it in practice, spot recurring friction, decide what is worth changing, and evolve the harness based on what you learn. You’ll capture those decisions in an HDR and put the updated harness back to work to see whether it actually performs better.
By the end of the day, you’ll have built an agent harness, created a continuous-improvement loop around it, and documented your first improvement as an HDR: Harness Decision Record. Not just a better harness, but one that can keep getting better.
Your AI Coding Assistant Customizations Have Zero Tests. Let's Fix That.
You'd never ship a function without a test. But the instruction files and agent definitions shaping every AI response in your codebase? Those get a code review where the best anyone can do is read the markdown and hope.
Teams are shipping these customizations alongside every feature, and reviewers are doing their best. But there's nothing to run. A misconfigured instruction file won't fail a build or throw an exception. It surfaces later, when a prompt returns output that's subtly, consequentially wrong. Same prompt, different customizations, wildly different results. It's "works on my machine" for AI.
This talk introduces practical evaluation methods for the full range of AI assistant customizations — instruction files, agent definitions, skills, and rules. You'll learn how to define what "correct" means for an instruction file or skill, how to choose between deterministic checks, LLM-as-judge rubrics, and human review, and how to wire evaluations into your existing PR workflow — so customizations face the same quality bar as the rest of your codebase.
You'll leave with a concrete framework for making AI customizations something your team can verify, not just eyeball.
GenAI from the trenches: from legacy to a modern development stack
Transforming legacy code into a modern software stack has always been a complex challenge, but generative AI is redefining how we approach this work. In this talk, we share how we built an AI-powered conversion pipeline to transform a legacy low-code solution into a modern Vue and TypeScript stack, achieving an 85 percent reduction in manual effort and greatly accelerating the modernization process.
Our story focuses on the specific challenges we encountered while building this conversion pipeline and the strategies we used to overcome them. From managing intricate legacy structures to ensuring the correctness of AI-generated code, we’ll walk you through the iterative steps we took to refine and validate our approach. You’ll see how we used GPT-4 to continuously improve prompt precision, and how we leveraged a Human-in-the-Loop framework to keep control over quality and reliability.
This session provides practical, actionable insights into prompt engineering, pipeline design, and effectively integrating generative AI as a development tool. You’ll learn how AI can streamline tedious tasks and tackle complex modernization challenges, yet remains a convenience rather than a replacement for specialized, developer-led work. Our main takeaway is clear: AI is a powerful collaborator that empowers developers to elevate their impact, keeping them firmly in command of the process.
Zero to Production: Rapid Prototyping with AI Tools
This fast-paced, hands-on session demonstrates how modern AI tools can dramatically accelerate the software development lifecycle, enabling practitioners to go from concept to production in record time. We'll explore a curated selection of cutting-edge AI-powered development tools, highlighting their unique strengths and optimal use cases in the prototyping workflow. Through practical demonstration, attendees will witness the creation and deployment of a fully functional application in just 50 minutes. The session showcases how strategic integration of these AI tools can eliminate traditional development bottlenecks, streamline collaboration, and transform the way teams build software. Whether you're a developer, designer, or tech enthusiast, this talk offers actionable insights and a clear roadmap for leveraging AI to turn ideas into working products with unprecedented velocity.
AI at Every Step: Why Are You Still Doing the Dev Cycle by Hand?
As AI agents become increasingly capable, the challenge for software teams shifts from using AI to orchestrating it responsibly.
This talk explores how AI Augmented Engineering can be applied across multiple phases of software development — from refinement to review — by combining Context Engineering with purpose-built AI agents. Each agent operates with a clear responsibility, explicit boundaries, and structured hand-offs.
We open the hood on real-world implementations: how agents are configured, how context is preserved between steps, and how humans stay in the loop as gatekeepers rather than passive observers.
The result is a pragmatic alternative to "vibe coding": a controllable, auditable, and scalable way of integrating AI into professional software engineering.
Agents Don't Sleep: How AI Turns a 3 AM Incident into a Reviewable Fix
It’s 3 a.m. A deployment from earlier that evening introduced a subtle bug and a critical service goes down. Your pager goes off. But by the time you open your laptop, an AI agent has already triaged the incident, traced the root cause back to a specific code change, and opened a pull request with a fix. Your job: read the analysis, review the code, and hit merge.
In this session, I show how AI agents can take over the operational side of DevOps. Unlike development agents, which work step by step with human approval, ops agents need to act autonomously: detecting anomalies, investigating root causes, and preparing fixes while no one is watching. You’ll see how to design an agent pipeline that monitors your production environment, performs automated triage when something breaks, and presents both a root cause analysis and a code fix for human review.
You’ll walk away with a concrete blueprint for agent-driven incident response and a realistic perspective on what you can deploy today versus what is still emerging.
The Harness Is Never Done: Continuous Improvement for Agentic Systems
We’ve learned to continuously improve our software. We fix bugs, run retros, review incidents, and turn recurring problems into changes. But when it comes to the systems we build around our agents, we often treat them as finished once they work.
Until the same annoyance happens for the fifth time.
You ask for an integration test and get a unit test. You correct the agent, adjust an instruction, and move on. A week later, you have the same conversation again. The harness works, but it isn’t getting better.
This talk explores what happens when we apply the principles of continuous improvement to the agent harness itself. Instead of treating repeated friction as something to work around, we’ll look at how those interactions can become signals that tell us where the harness needs to change.
Along the way, we’ll explore practical ways to discover these signals, turn them into deliberate improvements, and create a feedback loop between how we use our agents and how we build the systems around them.
Because the goal isn't to write the perfect harness. It’s to build one that gets better as you use it.
NDC London 2027 Sessionize Event Upcoming
Bitbash 2027 Sessionize Event Upcoming
Techorama 2026 Netherlands Sessionize Event Upcoming
ISKS 2026 - Built to last Sessionize Event
Future Tech 2026 Sessionize Event
Bitbash 2026 Sessionize Event
AI Lowlands 2025 Sessionize Event
techcamp 2025 Sessionize Event
Bitbash 2025 Sessionize Event
The Next Web 2024
GenAI from the trenches: from legacy to a modern development stack
AI Community Day Sessionize Event
Please note that Sessionize is not responsible for the accuracy or validity of the data provided by speakers. If you suspect this profile to be fake or spam, please let us know.
Jump to top