XConf APAC | Singapore

If AI can build software, then what's a software engineer's role? What do software engineer actually do?

This talk will explore that very question. By examining the fundamental similarities and differences between traditional software development by 100% human effort compared to AI-assisted creation, we will notice the timeless principle and identify the timeless skills that will empower software engineers to thrive in an AI-driven future.

AI is reshaping how we write code, but code generation is not software delivery. As teams rush to adopt AI tools, they often fall into hidden traps. This talk explores the most common pitfalls in AI-assisted development, and offers practical ways to pair with AI responsibly.

The software delivery lifecycle is changing faster than ever. Generative AI tools promise to accelerate every stage — from ideation and design to coding, QA, and release — but most teams are still figuring out how to integrate them without breaking quality, culture, or compliance.

In this talk, we’ll explore practical patterns for building an AI-augmented SDLC that drives developer effectiveness and organisational resilience. Learn how to use GenAI to get a sustainable engineering advantage.

AI in software engineering has generated an enormous amount of attention, with online stories ranging from spectacular successes to cautionary failures. But what’s really happening in practice when AI meets the day-to-day work of coding?

This panel will dig into lived experiences of applying AI within the code development phase. Building on earlier talks around AI in software engineering, the discussion will uncover what’s working, what’s not, and where the biggest opportunities and risks lie.

Turning LLM-enabled prototypes into reliable production systems is a thorny challenge. This talk examines these challenges from the perspectives of data science and cybersecurity, offering mental models to both articulate LLM risks and highlight useful applications. However you use LLMs in your stack, this talk has something for you.

Large language models (LLMs) generate diverse responses, yet trial-and-error prompt engineering is inefficient for finding the ideal ones. What other levers exist? This talk introduces a principled approach to steering word sampling during generation, giving users and developers greater control to align outputs with tasks, goals and business needs.

This talk explores why chatbot testing is different from traditional QA, covering the test lifecycle (NLU, flows, integration), real world challenges like flakiness and data variety, and lessons learned from practice. As conversational AI grows, testing must evolve – join us to discuss and help shape this emerging practice.

Discover how we helped a recruitment SaaS company use AI to compress the product discovery cycle from weeks to days. In this talk, you’ll see how rapid prototyping, user validation, and the FOREST AI Readiness Framework helped unlock innovation. Learn how AI can accelerate your product teams while aligning with strategic and customer goals.

AI in software delivery is often showcased through “vibe coding” stories — models that generate code from vague prompts. But delivery is much more than just writing code. This panel explores how AI is beginning to transform the often-overlooked stages of delivery: from design and product thinking to testing, security, and operations.

Building on earlier talks about security and product design, the discussion will push into new territory, asking: Where else can AI add real value in the delivery lifecycle? What are the opportunities, risks, and implications for teams?