Researcher, builder, communicator

About Pete

I work across language models, evaluation, systems-building, and technical communication.

I work on making language models more reliable in practice, especially through behavior shaping, reasoning, evaluation, and the engineering work needed to turn research into usable systems.

My recent work concentrates on prompting, context engineering, reasoning, and post-training, especially when these techniques need to hold up outside controlled demos.

Behavior shapingReasoning modelsEvaluation and reliabilitySingapore

At a glance

The fastest way to understand what I do, how I work, and where I am operating from.

Current role

Incoming PhD student @ NTU Singapore

Preparing for PhD research across reasoning models, evaluation, and reliable AI systems.

Working style

Research to systems

I like moving between ideas, experiments, implementation, and communication.

Base

Singapore

Working with regional and international collaborators from Singapore.

What drives my work

The main themes connecting the papers, models, tools, and talks.

Behavior shaping

Studying how model behavior changes under context engineering, prompting, supervised fine-tuning, and reinforcement fine-tuning.

Reasoning models

Building and analyzing open reasoning models, including Thai and medical settings where reliability matters.

Evaluation and reliability

Designing evaluations for structured outputs, failure analysis, robustness, and practical deployment risk.

Applied systems

Turning research into usable tools, experiments, and prototypes for real teams and domain experts.

What I am working on now

The current work mixes language-model research with the engineering and communication needed to make that work useful to other people.

Current focus

Studying post-training and reasoning behavior in open language models.
Investigating inference-time strategies, context engineering, and agentic workflows.
Working on medical reasoning and robustness in collaboration with domain partners.
Turning current research into talks, notes, and tooling that other teams can reuse.

How I tend to collaborate

I work with academic, clinical, and industry partners on reasoning, prompting, and domain-specific applications.
I like projects that connect research questions to deployable prototypes, datasets, and evaluation workflows.
My default mode is moving between ideas, experiments, implementation, and communication instead of treating them as separate jobs.

How I communicate the work

I treat public explanation as part of the work, not as an afterthought once the work is done.
I regularly turn experiments into talks, blog posts, internal research sessions, and reusable notes.
I like writing while I learn because it tests whether an idea is clear enough to teach and robust enough to reuse.

Background

The longer arc behind the current research and engineering work.

What I study

I am interested in how model behavior changes under human preferences, task constraints, domain requirements, and cultural context.

How I think

My recent work concentrates on prompting, context engineering, reasoning, and post-training, especially when these techniques need to hold up outside controlled demos.

Why it matters

Before focusing full-time on research, I spent years building software across mobile, web, data, and developer tooling. That background still shapes how I evaluate ideas: I care about whether they can be implemented, tested, and explained clearly.

Selected milestones

A short view of the education, work, and research moments that shaped how I approach language models and systems-building.

Spotlight milestones

2026 — Incoming PhD student, Nanyang Technological University (NTU Singapore)

Preparing to begin doctoral research on reasoning models, evaluation, and reliable AI systems.
Feb 2026 – May 2026 — Research Scientist, Typhoon team (SCB DataX R&I)

Advised fellow research scientists on language model research, supported internal knowledge sharing, and helped connect agentic AI work to product and strategy decisions.
Oct 2024 – Jan 2026 — Research Scientist, Typhoon team (SCB 10X R&D)

Conducted research on LLM reasoning and post-training, led work on Typhoon T1, and helped drive external research collaborations.
Sep 2024 — M.Eng., Ritsumeikan University (Valedictorian)

Graduated as valedictorian (4.93/5.00) with a thesis on null-shot prompting and hallucination in large language models.
2022 — Research Associate (IC2, SIT KMUTT)

Led end-to-end machine learning, AR, and web projects; presented Krathu-500; and organized a Kaggle competition.
Feb 2022 — B.Sc. Computer Science, King Mongkut's University of Technology Thonburi (First-class honours)

Graduated with first-class honours (3.99/4.00). Senior project: COEUS, automatic English mind map generation from text.
2019–2020 — Siriraj Hospital — React Native developer

Built product features and data collection flows as part of the mobile development team.

Education

Valedictorian (4.93/5.00). Thesis: Null-Shot Prompting: Rethinking Prompting Large Language Models With Hallucination.

First-class honours (3.99/4.00). Senior project: COEUS, automatic English mind map generation from text.

Read the extended timeline

2026 — Incoming PhD student, Nanyang Technological University (NTU Singapore)

Preparing to begin doctoral research on reasoning models, evaluation, and reliable AI systems.
Feb 2026 – May 2026 — Research Scientist, Typhoon team (SCB DataX R&I)

Advised fellow research scientists on language model research, supported internal knowledge sharing, and helped connect agentic AI work to product and strategy decisions.
2025 — Typhoon Application Week

Built and shipped more than seven LLM-powered applications in a rapid prototyping sprint.
2025 — Typhoon T1 (research preview)

Contributed to the research preview of Typhoon T1, the first open reasoning model developed for Southeast Asia.
Oct 2024 – Jan 2026 — Research Scientist, Typhoon team (SCB 10X R&D)

Conducted research on LLM reasoning and post-training, led work on Typhoon T1, and helped drive external research collaborations.
Sep 2024 — M.Eng., Ritsumeikan University (Valedictorian)

Graduated as valedictorian (4.93/5.00) with a thesis on null-shot prompting and hallucination in large language models.
2022 — Ethical Hacking Workshop

Completed a five-day red-team training workshop organized by NIA and NBTC.
2022 — Research Associate (IC2, SIT KMUTT)

Led end-to-end machine learning, AR, and web projects; presented Krathu-500; and organized a Kaggle competition.
Feb 2022 — B.Sc. Computer Science, King Mongkut's University of Technology Thonburi (First-class honours)

Graduated with first-class honours (3.99/4.00). Senior project: COEUS, automatic English mind map generation from text.
2021 — LINE Thailand Internship

Worked on LINE Shopping as a full-stack intern, including Vue 2-to-3 migration work and microservice-related tasks.
2020–2021 — Siriraj Hospital (Mahidol) — Tech lead and Flutter developer

Led Flutter development, helped migrate the app from React Native to Flutter, and improved app size and loading performance.
2020 — COVID‑19 Support Platform (KMUTT)

Led a 10-person team building a COVID-19 support platform that tracked more than 300 shipments and served more than 50 hospitals.
2020 — Huawei ICT APAC (Network Track)

Participated in the regional Huawei ICT competition with a focus on routing and switching.
2019–2020 — Siriraj Hospital — React Native developer

Built product features and data collection flows as part of the mobile development team.
2019 — Google Tech Challenge (Bangkok)

Competed as part of a selected team in a challenge spanning both coding and non-coding problem solving.
2019 — Young Webmaster Camp 17th

Selected for the programming track and built an MVP for a social-impact project.