Talk Abstracts & Bio

This page contains abstracts for my talks and my speaker bio.

Bio

English

Sun Seng David TAN (a.k.a Sunix) is an experienced software engineer and tech lead with over 15 years of expertise in developing Java applications and cloud-native solutions. Formerly at Red Hat, he has contributed extensively to major open-source projects such as Eclipse JKube, Fabric8 Kubernetes Java Client, Eclipse Che, and Eclipse Theia. As a passionate advocate for the Java community, Sun served as the ParisJUG Leader from 2019 to 2023 and has been an active crew member since 2015. He also supported Devoxx France as part of the “gilets rouges” team, playing a key role in organizing one of Europe’s largest developer events.

In 2024, Sun joined Sciam as a tech lead consultant, bringing his expertise in guiding companies toward innovative and high-performing solutions. Passionate about continuous learning and knowledge sharing, Sun regularly speaks at conferences and remains involved in numerous open-source projects.

In his spare time, Sun brews his own beer in his kitchen, combining his passion for technology and artisanal creation.

Français

Sun est un ingĂ©nieur logiciel et tech lead expĂ©rimentĂ© avec plus de 15 ans d’expertise dans le dĂ©veloppement d’applications Java et de solutions cloud-native. Anciennement chez Red Hat, il a contribuĂ© Ă  des projets open-source majeurs tels que Eclipse JKube, Fabric8 Kubernetes Java Client, Eclipse Che et Eclipse Theia. Actif dans la communautĂ© Java, Sun fait partie de l’équipe ParisJUG depuis 2015 et a occupĂ© le rĂŽle de JUG Leader de 2019 Ă  2023.

En 2024, Sun rejoint Sciam en tant que consultant tech lead, apportant son expertise dans l’accompagnement des entreprises vers des solutions innovantes et performantes. PassionnĂ© par l’apprentissage continu et le partage de connaissances, Sun intervient rĂ©guliĂšrement lors de confĂ©rences et reste impliquĂ© dans de nombreux projets open-source.

Pendant son temps libre, Sun brasse sa propre biÚre dans sa cuisine, combinant sa passion pour la technologie et la création artisanale.

Talk Abstracts

GI DD: GitHub Issue Driven Development

Title: How I Finished My First Side Project Without Coding Thanks to GitHub Copilot

You might say, yet again another talk about vibe coding and AI, you may be right but it is slightly different here. Like many developers, my life is littered with unfinished side projects. But this time was different—I actually finished, and I did it almost entirely without opening an IDE. This talk explores a revolutionary approach I call “GitHub Issue Driven Development” (GI DD), where GitHub Copilot becomes your coding companion in the most literal sense.

In this session, I’ll share my journey of building a Progressive Web App for managing gift cards using GitHub Copilot’s workspace feature. Instead of traditional coding sessions, I created issues, assigned them to @copilot, and watched as it generated code, opened pull requests, and iterated based on my feedback. The workflow was simple: write clear issue descriptions, let Copilot handle the implementation, review the PR, provide feedback through comments, and merge.

What you’ll learn:

  • How to effectively delegate entire features to GitHub Copilot through GitHub Issues
  • The iterative workflow of creating issues, reviewing PRs, and providing feedback using @copilot mentions
  • Real-world examples of where AI excels and where human guidance is crucial (spoiler: barcode generation libraries!)
  • How to break down projects into manageable issues that AI can handle
  • The visibility and transparency benefits of seeing Copilot’s “thinking” process through session logs and screenshots
  • Practical tips for writing effective prompts and managing the AI collaboration process

Key takeaways: The barrier to finishing side projects isn’t just time or skill—it’s about having the right collaboration model. This talk demonstrates how pairing with AI can transform your approach from “I need a weekend to code this” to “I can make progress during coffee breaks.” Whether you’re skeptical about AI-assisted development or eager to optimize your workflow, you’ll leave with practical strategies for actually shipping your next side project.

This presentation may also explore how AI agents could further enhance GitHub Copilot’s capabilities, making the development process even more seamless and efficient.

Based on: Building a Gift Card Management App with GitHub Copilot


GI DD: GitHub Issue Driven Development (Version Française)

Titre: Comment j’ai terminĂ© mon premier side project sans coder grĂące Ă  GitHub Copilot

Vous vous dites peut-ĂȘtre, encore une prĂ©sentation sur le “vibe coding” et l’IA, vous avez peut-ĂȘtre raison mais c’est un peu diffĂ©rent ici. Comme beaucoup de dĂ©veloppeurs, ma vie est remplie de side projects inachevĂ©s. Mais cette fois, c’était diffĂ©rent—j’ai enfin terminĂ© un side project, et je l’ai fait presque entiĂšrement sans ouvrir un IDE. Cette prĂ©sentation explore une approche nouvelle que j’appelle “GitHub Issue Driven Development” (GI DD), oĂč GitHub Copilot couplĂ© Ă  Github Issue m’a permis de coder une application complĂšte 
 pendant mes pauses cafĂ© ;)

Dans cette session, je partagerai mon expĂ©rience dans le dĂ©veloppement d’une Progressive Web App pour gĂ©rer des cartes-cadeaux. Au lieu de sessions de codage traditionnelles, j’ai créé des issues, les ai assignĂ©es Ă  @copilot, et j’ai regardĂ© le code se gĂ©nĂ©rer, les pull requests s’ouvrir, et les itĂ©rations se faire en fonction de mes retours. Et enfin merger lorsque j’étais satisfait !

Ce que vous apprendrez :

  • Comment dĂ©lĂ©guer efficacement des fonctionnalitĂ©s entiĂšres Ă  GitHub Copilot via les GitHub Issues
  • Le workflow itĂ©ratif de crĂ©ation d’issues, rĂ©vision de PRs, et retours via les mentions @copilot
  • Des exemples concrets oĂč l’IA excelle et oĂč l’orientation humaine est cruciale (spoiler : les bibliothĂšques de gĂ©nĂ©ration de codes-barres !)
  • Comment dĂ©composer des projets en issues gĂ©rables que l’IA peut traiter
  • Les avantages de visibilitĂ© et de transparence en voyant le processus de “rĂ©flexion” de Copilot via les logs de session et captures d’écran
  • Conseils pratiques pour Ă©crire des prompts efficaces et gĂ©rer le processus de collaboration avec l’IA

Points clĂ©s Ă  retenir : Cette prĂ©sentation dĂ©montre comment le binĂŽmage avec l’IA peut transformer votre approche de “j’ai besoin de 5 week-ends entiers et 3 soirĂ©es par semaine pour coder cette application” Ă  “je peux demander Ă  copilot de travailler dessus pendant mes pauses cafĂ©â€. Que vous soyez sceptique vis-Ă -vis du dĂ©veloppement assistĂ© par IA ou dĂ©sireux d’optimiser votre workflow, vous repartirez avec des stratĂ©gies pratiques pour enfin terminer votre prochain side project.

Dans cette prĂ©sentation, nous explorerons Ă©galement l’utilisation personnalisĂ©e des agents IA pour amĂ©liorer davantage les capacitĂ©s de GitHub Copilot, rendant le processus de dĂ©veloppement encore plus fluide et efficace.

Basé sur : Building a Gift Card Management App with GitHub Copilot