We build courses the way developers read.
Neuronest started with a simple observation: most AI learning material is either too shallow or too scattered. We set out to write something different.
Back to HomeAbout Neuronest
Neuronest was founded in Bangkok in 2021 by a small group of developers who had worked with machine learning systems in production and found that most available learning resources fell into one of two categories: introductory material that stopped well short of anything useful, or research-level writing that assumed background most working developers don't have.
We started writing our own notes. Over time, those notes became structured courses — complete with exercise sets, reference solutions, and worked notebooks. We opened them to other developers in 2022, starting with a small group of students in Bangkok and a few online participants from across Southeast Asia.
The school has grown since then, but the approach remains the same. We write long-form, readable material. We pair every technical section with a working code reference. We do not use video. We do not use slide decks. We write the way we would want to read — with enough detail to actually understand something, not just recognize it.
Our Mission
To produce learning material that working developers can use to build genuine understanding of AI systems — material that is specific enough to be useful and clear enough to read without a dictionary.
Our Students
Most of our students are software engineers or data analysts who want to move into AI work or deepen their understanding of systems they are already using. We do not require a mathematics background beyond what a working programmer typically has.
Our Location
We are based on Sukhumvit Soi 21 in Watthana, Bangkok. All courses run online. Students may arrange in-person study sessions at our Bangkok office by appointment.
The Team
Three people write and maintain all Neuronest course material.
Thanakorn Kittipong
Previously a senior ML engineer at a Bangkok-based logistics company. Wrote the Foundations course from scratch in 2021. Focuses on gradient methods and model evaluation.
Priya Subramaniam
Software engineer with a focus on NLP tooling and retrieval systems. Designed and maintains the Building With Open-Source Language Models course. Has worked with teams across Thailand and India.
Nattawut Wongkham
Researcher with a background in statistics and a long habit of reading machine learning papers. Moderates the monthly reading group and writes the pre-session notes distributed to participants.
Our Standards
How we maintain quality across every course we publish.
Written and Reviewed Internally
All course material is written by a Neuronest instructor and reviewed by at least one other team member before publication. We do not commission external writers.
Every Notebook Runs
We run every example notebook against a clean Python environment before each course intake. If something breaks, we fix it before students receive access.
Regular Content Updates
The field moves quickly. We review course content twice per year and update sections where the standard approach or available libraries have changed meaningfully.
Student Data Privacy
We store only the data needed to manage course access and communicate with enrolled students. We follow Thai PDPA requirements and do not share student information with third parties.
Forum Moderation
Discussion threads are monitored by course instructors. We respond to questions on working days and keep threads on-topic and useful as searchable references.
Honest Scope Descriptions
We describe what each course covers and what it does not. We do not inflate difficulty or outcome claims. Students should know what they are signing up for before they pay.
What We Value in Technical Education
Most technical education exists on a spectrum between two failure modes. On one end, there is material that is accessible but shallow — it gives you a vocabulary without the understanding needed to use it. On the other, there is material written primarily for researchers — precise and complete, but opaque to anyone who has not spent years reading the same literature.
Neuronest aims for neither. We write for developers who are comfortable with code, who can read an equation if it is explained carefully, and who want to understand what their tools are actually doing. We spend considerable time on worked examples, not because examples replace explanation, but because seeing a concept applied in code often clarifies what a written description left vague.
We also take the sequencing of material seriously. Machine learning involves a set of ideas that build on each other. Gradient descent is hard to explain meaningfully without a clear picture of what loss functions are doing. Retrieval-augmented generation is confusing without first understanding what a language model does and does not know. We try to present concepts in the order that makes each one legible, not in the order that a syllabus committee might arrange them.
The reading group is a different kind of offering — less structured, more conversational. Papers are how the field actually communicates new ideas, and learning to read them is a skill that does not come automatically even to experienced practitioners. The group exists for people who want a regular, low-pressure context for developing that skill alongside others.
Talk to us before you decide.
Send a message and we will describe which course suits your current background and what the first few weeks look like in practice.