XConf APAC 2025 | Singapore
as the organisers put it
XConf APAC is back in Singapore on Oct 31. Join our annual tech conference created by technologists, for technologists. Register today!
| title | author | length |
|---|---|---|
| Singapore highlights | 01m | |
Highlights from XConf APAC 2025 show concrete techniques and outcomes across the software delivery life cycle with a focus on practical tooling and workflows… | ||
| Panel Discussion: Beyond vibe coding: Expanding AI’s role in delivery | Esther Tham, Donald Ong, Lilly Ryan, Katie Martin | 46m |
Panel argues AI speeds feedback and prototyping (a 5‑day proof‑of‑concept example) but does not replace experience: teams embed continuous embedded security… | ||
| Prototyping at the speed of thought: AI as a discovery accelerator | Katie Martin | 25m |
A recruitment SaaS use case goes from messy problem to user‑tested solution under five business days by using Figma Make to generate a high‑fidelity prototype… | ||
| Chatbot testing in the real world: Making sense of an emerging practice | Tayakorn Gindavichu, Warintorn Phusomsai | 22m |
Claims that chatbot QA is about testing experience, not just correctness and that a custom NLU testing framework catches intent drift early; describes running… | ||
| Smarter word generation in LLMs: Guiding sampling through output probability distributions | Runyan Tan | 22m |
Introduces a decoding rule called Pillars that sets a dynamic cutoff equal to the sum of squared token probabilities so only tokens with p >= sum(p^2) are… | ||
| The Sorcerer’s Apprentice: The craft of building reliable solutions with LLMs | Lilly Ryan, Ned Letcher | 32m |
LLM prototypes often fail on real data because models are non‑deterministic, can’t count, and have short corruptible memory; reliability requires pruning… | ||
| Panel discussion: AI in code development – hype vs reality | Thao Dang, Jia Yi GOH, Aditi Agarwal, Chris Chakrit Riddhagni, Wenbo Fan | 41m |
Panel reports that AI is adopted widely but usually brings modest gains: 90% adoption, 72% report only moderate productivity uplift. Practical wins include… | ||
| AI-augmented SDLC: Rethinking how we deliver software | Aditi Agarwal | 26m |
Generative AI capability doubles roughly every seven months but organisations report only a 5% productivity impact because tools amplify both strengths and… | ||
| The new gravity of software engineering in the AI-first era | Sarah Taraporewalla, May Xu | 27m |
Claim: AI reshapes the “physics” of software so design, testing and delivery compress into near‑instant loops; prototypes and rewrites (LangChain 1.0 rewrote a… | ||
| AI-assisted development: Common pitfalls and smarter practices | Wenbo Fan | 39m |
A Stanford study shows AI code generation raises raw output by 30–40% while rework and refactors jump to 20–30%, so net productivity gains stay limited. Real… | ||
| Beyond the code: Rediscovering the timeless skills of software engineering in the AI era | Chris Chakrit Riddhagni | 28m |
AI coding demos (Codex / "C code", Cursor, cloud code) produce runnable pages but fail when specs grow: large context makes LMs lose focus — context… | ||
| nothing matches | ||