PyCon APAC 2025 in Manila Highlights - AI, GIL, and Inclusivity

The 1st part of a 2-part series of my experience attending the final PyCon APAC in Manila

Introduction

Happy to be back in Manila! I was in Manila last year for PyCon PH and this year I was also invited to be a panelist in one of the panel discussion about AI (what else, right? 😜).

PyCon APAC 2025 in Manila was an incredible experience, both personally and professionally. As the CEO of LaLoka Labs, I was there to promote Kafkai, our AI-powered content strategy and generation tool. But beyond that, I was also attending as a director of the Python Asia Organization, working to foster the Python community across the region. This post is the first of two parts covering my journey through PyCon APAC 2025, sharing insights, key takeaways, and memorable moments.

Kafkai at PyCon APAC: A Chance to Connect

I opted for a Corporate-level ticket to support the conference, which came with some perks, including the ability to distribute marketing materials. This was a golden opportunity to introduce Kafkai to a new audience.

We designed special flyers with an exclusive PyCon APAC discount, making sure attendees got the best deal possible. The response was enthusiastic, many developers and content creators stopped by to learn how Kafkai could help them streamline their work. It was particularly rewarding to see how AI-generated content resonated with the Python community, a group that thrives on automation and efficiency.

Kafkai marketing flyers and promotional materials laid out on a table at the PyCon APAC 2025 patron booth.

The GIL and Python 3.13: A Key Technical Session

One of the standout talks I attended was "An Introduction to the GIL for Python Beginners: Disabling It in Python 3.13 and Leveraging Concurrency." The session dove into the intricacies of the Global Interpreter Lock (GIL) and how its optional removal in Python 3.13 opens new doors for parallelism in Python.

The speaker for the talk was Yu Saito, a Technical Specialist for Azure at Microsoft Japan

A three-slide collage showing a PyCon APAC 2025 speaker presenting how to disable Python’s GIL, including instructions for different operating systems, concluding notes on performance trade-offs, and comparisons with other concurrency methods.

Key takeaways:

Python 3.13 introduces python3.13t.exe, which allows true multi-threading.
The Global Interpreter Lock (GIL) has historically been a limitation for multi-threading in Python, but its removal in Python 3.13 opens new possibilities for performance improvements.
Disabling the GIL enables true parallel execution, allowing better resource utilization for CPU-bound tasks.
Developers must exercise caution when disabling the GIL, as it introduces potential risks such as race conditions and requires proper synchronization techniques.
Thread-safe libraries become crucial when working in an environment where the GIL is disabled.
Benchmarks have shown that removing the GIL can significantly boost performance in multi-threaded applications, though the extent of improvement varies by workload.
The Python community is actively exploring best practices and tools to help developers transition to GIL-free Python.
Multi-threading vs. multi-processing: GIL-free Python enables true parallelism.
While disabling GIL improves performance, it comes with risks: race conditions and the need for thread-safe libraries.
C/C++ extensions remain a viable way to bypass GIL restrictions in older Python versions.

For anyone working on CPU-bound Python applications, this is an exciting development. I left the session eager to experiment with the disabled GIL feature in future projects.

Diversity & Inclusion Panel: A Critical Discussion

One of the key sessions I attended at PyCon APAC 2025 was the Diversity & Inclusion panel discussion. As a long-time advocate for inclusivity in tech, I found it insightful to hear perspectives from industry leaders on fostering a more diverse and welcoming Python community.

The panelist for the panel was:

Maaya Ishida, organizer for PyLadies Tokyo and one of the board member for PyCon JP Association
Kalyan Prasad, one of the Chairs for PyConf Hyderabad 2025
Aryn Choong, a Software Consultant and Manager of for startups.
Leonora Sision, who manages and mentors a diverse Team of Manual Testers and Automation Engineers from different countries
and Ivy Fung, committee member of PyCon MY Pyladies KL, as the moderator.

A two-image collage showing five panelists standing together on stage in the top photo, and a panel discussion on fostering diversity and inclusion in the Python community in the bottom photo.

Key Discussion Points:

The importance of mentorship and accessibility for underrepresented groups.
How local Python communities can create more inclusive spaces.
Real-world initiatives that have successfully diversified open-source communities:
The implementation of mentorship programs that pair experienced developers with newcomers, providing them with the guidance and confidence to contribute.
The introduction of language translation efforts to make documentation and learning resources more accessible to non-English speakers.
The establishment of financial aid and grant programs to support contributors from underrepresented regions in attending PyCon and other Python-related conferences.
Community-driven efforts to ensure that event speakers and leadership teams are more representative of the broader community.
Case studies of open-source projects that have successfully implemented inclusive policies, leading to increased engagement from diverse contributors.

While the panel discussion itself didn’t directly cover this, it sparked ideas for me on how organizations like LaLoka Labs and our products like Kafkai can contribute to making the tech ecosystem more equitable. Ensuring AI tools are accessible to a diverse range of developers and customers is a crucial step in fostering inclusivity. AI is a powerful tool to bridge gaps, whether those are gender or economic gaps between individuals or businesses. Hearing the different perspectives from panelists and audience members reinforced the notion that inclusivity is an ongoing effort requiring active participation from all of us.

A Dynamic and Engaging Community

Beyond the sessions, the heart of PyCon APAC was the community. I had insightful conversations with Python developers from across Asia, many of whom shared their experiences working with AI, automation, and content generation. As a director of the Python Asia Organization, it was fulfilling to see how our regional Python communities are growing and collaborating more than ever.

Stay tuned for Part 2, where I’ll cover my role as a speaker, the Future of Software Engineering in the Age of AI panel, and the overall impact of PyCon APAC 2025 on the Python ecosystem.