Aaron Soh Jun Qi and Ian Tan Xeng, freshmen in Nanyang Technological University's (NTU) College of Computing and Data Science (CCDS) Data Science and Artificial Intelligence programme, have made headlines by securing spots among just 22 undergraduates worldwide selected for the prestigious AAAI-26 Undergraduate Consortium. This achievement underscores NTU's growing prowess in nurturing undergraduate AI research talent right from the early years of university.

Hosted alongside the 40th AAAI Conference on Artificial Intelligence in Singapore from January 20 to 27, 2026—the first time the event was held outside North America—the Undergraduate Consortium provided these young researchers with a platform to present their innovative proposals to global AI leaders. Their participation highlights how Singapore's universities are positioning themselves at the forefront of artificial intelligence advancement.

Understanding the AAAI-26 Undergraduate Consortium

The Association for the Advancement of Artificial Intelligence (AAAI) Undergraduate Consortium (UC) is a selective two-day programme designed specifically for promising undergraduates embarking on AI research careers. Held on January 20-21, 2026, at the Singapore EXPO, it drew nearly 200 applications globally, with only 22 scholars chosen based on a rigorous review of their two-page research proposals.

Participants engaged in tutorials, keynote lectures by experts like He Yang from A*STAR's Centre for Frontier AI Research and Lizi Liao from Singapore Management University, panel discussions, science communication workshops, oral presentations (five minutes each plus Q&A), and poster sessions. All accepted students received full travel funding, enabling focus on learning and networking. This immersive experience not only refines research ideas but also connects budding researchers with mentors from academia and industry, fostering the next generation of AI innovators.

Aaron Soh Jun Qi's Groundbreaking Proposal

Aaron Soh Jun Qi's research focuses on bridging speech processing technologies with the textual strengths of large language models (LLMs). Initially exploring applications like assistive technologies for the hearing impaired and real-time transcriptions for meetings or lectures, Aaron pivoted to address a core challenge: the scarcity and high cost of annotated speech data needed for training models.

His proposal outlines novel methods to overcome this limitation, enabling more efficient and accessible speech-to-text systems. Under the guidance of NTU research fellow Wei Tianjun, Aaron emphasised the importance of a strong research foundation. As he noted, “My biggest takeaway under his guidance is that before we even begin running experiments, it is important to have a convincing abstract, background and introduction to justify and motivate your reasons for proposing something new in research.”

During the poster session, feedback from peers and experts helped Aaron identify flaws in his approach, spurring him to explore interdisciplinary AI applications further.

Ian Tan Xeng's Work on Efficient AI Fine-Tuning

Ian Tan Xeng tackled the stabilisation of training for Low-Rank Adapters (LoRA), a technique for efficiently fine-tuning large pre-trained models in resource-limited settings. LoRA reduces computational demands by updating only a small subset of parameters, but training instability remains a hurdle. Ian's proposal introduces methods to enhance reliability, drawing on recent studies that challenge conventional initialisation assumptions.

Mentored by Professor Kai Dong Feng from Yanshan University, who aided in refining experiments and theoretical justifications, Ian gained profound insights. “I was surprised by how much time, compute, and resources are required to actually conduct research,” he reflected. The conference reinforced the value of academic networks, motivating him to dive deeper into emerging AI areas.

The Transformative Conference Experience

Both students described the AAAI-26 UC as a whirlwind of inspiration. From Gardens by the Bay team-building dinners to interactions with Singapore-based AI luminaries, the event immersed them in the field's vibrancy. Presenting to an international audience honed their communication skills, while critiques sharpened their methodologies. As freshmen, this early exposure positions them ahead in NTU's competitive research landscape.

• Keynote insights on frontier AI from A*STAR and SMU experts.
• Poster feedback loops identifying real-world applicability gaps.
• Networking with 22 global peers, fostering potential collaborations.

This homegrown success, with AAAI-26 in Singapore, amplified national pride in local talent.Learn how to craft an academic CV for such opportunities.

Photo by Vitaly Gariev on Unsplash

Mentorship: The Backbone of Early Success

Mentorship proved pivotal. Wei Tianjun at NTU guided Aaron through structuring compelling proposals, stressing motivation before experimentation. Professor Feng supported Ian remotely, bolstering empirical and theoretical rigour. Such guidance exemplifies NTU CCDS's commitment to undergraduate involvement, where freshmen access labs and faculty early.

In Singapore's ecosystem, where AI ethics and compute resources are prioritised, mentors bridge academia-industry gaps. For aspiring researchers, seeking such support via research assistant jobs at universities like NTU is key.

NTU's College of Computing and Data Science in Focus

Launched to spearhead Singapore's AI push, NTU's CCDS offers the Bachelor of Computing (Honours) in Data Science and Artificial Intelligence—a four-year programme jointly with the School of Physical and Mathematical Sciences. It covers AI fundamentals (reinforcement learning, NLP, robotics), data analytics, ethics, and capstone projects tackling real challenges in healthcare and sustainability.

Unique features include hands-on industry immersions, Generative AI Lab access, and global exchanges. Graduates land roles at Google, Grab, and beyond, with strong starting salaries. CCDS fosters undergrad research via masterclasses and interdisciplinary institutes.Explore NTU CCDS.

Singapore's National AI Strategy and Talent Pipeline

Singapore's National AI Strategy 2.0 (NAIS 2.0), launched in 2023, commits billions to AI R&D, including a S$1 billion National AI Research and Development (NAIRD) Plan in 2026 for talent and public good applications. Universities like NTU (top 5-8 globally in AI rankings 2026) and NUS (top 7-11) drive this, with grads boasting 98% placement rates and S$80K starting pay.

Hosting AAAI-26 amplified this, coinciding with AI Research Week. NTU's Alan Turing AI Scholars programme expands to 60 annually by 2027.Read NAIS 2.0 details. For careers, check higher ed jobs in AI.

Future Prospects for NTU's AI Prodigies

Aaron plans lab-based research at NTU, venturing into multimodal AI. Ian aims to leverage connections for advanced projects. Their AAAI exposure equips them for publications, internships, and grad school. In Singapore's booming AI sector—projected to add thousands of jobs—these students epitomise the talent pipeline.

Implications for Undergraduate AI Research in Singapore

This feat signals shifting paradigms: freshmen competing globally, thanks to supportive curricula and national backing. Challenges like compute access persist, but solutions via shared resources and mentorship abound. Other unis like NUS mirror this, with similar AAAI successes.

Photo by Chunjiang on Unsplash

• Early research boosts PhD admissions and industry readiness.
• Ethical AI focus aligns with NAIS governance pillars.
• Interdisciplinary approaches address real-world impacts.

Actionable Insights for Aspiring AI Students

To emulate Aaron and Ian: Start with strong maths foundations, seek mentors early, craft motivated proposals, and attend conferences. NTU's programme emphasises responsible AI—vital amid rapid advancements. Explore scholarships and career advice for pathways.

For faculty and jobs, visit university jobs, higher ed jobs, rate my professor, and higher ed career advice.