PyConSG Education Summit 2025

Opening Remarks

Mr Gi - Retyree

Round table self-intro

Teaching Computer Science with Generative AI

Dr Tan Chee Wei - NTU

Data structures and algorithms are foundational topics in any computer science curriculum, yet students often find them abstract and challenging. To address this, together with my student Lim Jia Earn, we have developed a learner-centric platform, MyCodeWeapon, which integrates generative AI - including LLM agents and AI copilots for code - to support scalable, personalized learning.

How Informatics and AI is taught in South Korea high schools

Mr Giyeon Bang - South Korea high school computing teacher

South Korea is a global technology powerhouse with well known household names such as Samsung, SK, Naver, LG, Hyundai, etc. and is ranked as one of the world's top digitally connected society and highly ranked education system. How did its mainstream high schools lay a firm foundation in informatics (computer science) and AI literacy? We explore this from a high school computing teacher's perspective.

Bridging Bebras to National Olympiad in Informatics (NOI) and beyond

Ms Sophie Koh - Excel League

Bebras is the international benchmark for computational thinking and informatics, which does not involve programming. The National Olympiad in Informatics is a competitve programming contest which focuses on data structures and algorithms (and which NUS typically recommends C++). We have provided problem sets involving computational thinking and Python coding to improve accessibility to Computing students/learners which typically learn Python in schools.

PCDIT Chatbot for Helping Novice Programmers Write Code

Dr. Oka Kurniawan - SUTD

Generative AI has sparked a discussion on whether we still need to learn coding. One of the purpose of learning coding is to enhance problem solving skills and thinking. Novice programmers need help in their problem solving skills and PCDIT framework helps novice programmers to navigate the various barriers in problem solving through concrete steps. In this talk, we will share the experience of designing and deploying PCDIT chatbot which is meant to facilitate students in their problem solving when doing programming problems.

Exploring Multi-Agent Generative AI in Education and Career Advisory

Dr Yeo Wee Kiang - NUS

Students often face difficulties navigating vast resources with little personalised support. The AI-based solution combines large language models (LLMs) and multi-agent systems to deliver tailored educational and career advisory. The system's Tutor Section offers progressive teaching, contextual Q&A, and personalised review; its Career Advisory Section handles interview questions, skill gaps, cover letters, and resource recommendations.

PyLadiesSG Revived!

Ms Tay Hwee Shan - PyLadiesSG

After a decade of passing through different hands, PyLadiesSG is revived again this year, stronger and better than before! Get a glimpse of what we have accomplished so far and what to expect in the near future, and how you can contribute to PyLadiesSG regardless of your gender and background!

Un-Lightning Talks

1. Ashish Deepak Dandekar

Scaffolding the Problem-Solving Process for Introductory Computing Students: • Problem Context. Overview of traditional CS1 approaches and challenges faced by students leading to high failure rates. • Intervention. Proposed structured framework to provide scaffolding for students solve problems. • Impact. Implementation in a CS1 course as well as in an intensive clinic for failing students.

2. Lim Jingyi

How might we Inject Fun and Purpose when Teaching Coding? • Is there something that you wish could be created or automated with technology? • Let's explore with Python, OpenCV, APIs, electronics or other mashups. • Would you be able to beat the high score when Python challenges Google Dinosaur? • Or would you build a mini self-Driving Car that follows a road that you draw? • Or will you crawl the Singapore Map to help your parents locate ideal options for their next dream home? • Come ready to have fun, sweat, and contribute ideas.

3. Anand Subramaniam

Pythonic Grading that Rewards Sharing: • Flip the Plagiarism Narrative. Bonus marks trump penalties. Copying is a learning accelerant. • Python in the Trenches. Tokenize, shingle, and MinHash + Jaccard for rapid plagiarism detection. • Live Data Autopsy. Visual clusters of copies show who copied whom, when, and maybe why. • Four “Rules of Copying”. Evidence-based strategies students actually follow (and should).

AISG programmes

Dr Chen Weiqiang - AI Singapore

AI Apprenticeship programme (AIAP) • AI Apprenticeship Programme for Industry (AIAP(I)) (new!)

Teaching Goals and Plans - How we might help students improve problem-solving

Dr Norman Lee - SUTD

Often students are expected to pick up basic problem-solving skills by inferring strategies from solving stereotypical problems e.g. determining if an integer is a prime number, finding the sum of a series, finding the maximum in a list etc.

Make Learning Easy and Fun @ NLB LearnX

Mr Goh Soon Seng - NLB LearnX, PSF Fellow

We run quarterly workshops for beginners with no prior coding experience or knowledge of the Python Language. Using the Raspberry Pi Pico as a platform to teach, we endeavour to make learning to code in Python easy to grasp and enjoyable.
Participants are taught to build simple circuits using components such as buttons, LEDs, buzzers and bring these circuits to life using Python code. This interactive method provides immediate visual feedback. We also demonstrate that the pieces of knowledge gained can to put together to solve real-world problems like a pedestrian crossing/traffic light system or a burglar alarm system.
In our recent sessions, we have included an AI segment, where participants are introduced to tools like ChatGPT. We showed them how AI tools can be used, not only to assist them with writing, explaining, and debugging code, they can also use them to further their Python learning journey on their own.

Prof Leong Hon Wai - ex-NUS

Round Table Discussion

Topics to be submitted and voted on by participants on the day of the summit