About Synaptiq

Built around the problems developers actually face

Synaptiq started because the gap between a tutorial and a working model is larger than most online resources admit. We design courses that sit in that gap honestly.

← Back to Home

Our Story

How Synaptiq came together

Synaptiq was set up in Bangkok in 2022 by a small group of engineers who had spent time working on computer vision and NLP pipelines in production. The premise was simple: the materials available for learning practical deep learning were either too shallow — moving fast, skipping the mathematics — or too academic, disconnected from the kinds of decisions that come up when you are actually building something.

The school began with a single cohort course and a handful of students. Most were developers with Python experience who had hit a wall reading papers or following video tutorials that never quite connected to real implementation. The early sessions were spent figuring out exactly where that wall stood.

Over time the curriculum developed into what it is now: a self-paced foundations course, a ten-week cohort covering computer vision with open pretrained models, and quarterly office sessions for students who want structured input on their own projects. Every piece of content is tied to a runnable notebook.

Mission

To offer AI development education that is technically honest — one that does not paper over the mathematical foundations or pretend that "just use a pretrained model" covers everything worth knowing.

"The goal is not to fill seats. The goal is for students to leave with something that holds up — code that runs, understanding that transfers."

— Synaptiq founding notes, 2022

Values

Honesty about difficulty. We do not simplify past the point where the simplification becomes misleading.
Small groups by design. Cohort caps exist so feedback is real, not templated.
Runnable over readable. Every concept ships with working code, not just prose explanation.
No certificate theater. The deliverable is understanding and demonstrable work, not a PDF.

The Team

People behind the courses

The facilitators at Synaptiq have engineering backgrounds — they have worked on real systems and bring that context into the sessions.

Kasem Pratheep

Lead Facilitator

Previously built computer vision pipelines for logistics sorting systems. Leads the Computer Vision cohort course and office sessions.

Sutida Lertchai

Curriculum Designer

Designed the Neural Network Fundamentals curriculum and notebooks. Research background in optimization and training dynamics.

Nattawut Wongsa

Technical Reviewer

Reviews all notebook content for correctness before each cohort. Background in ML infrastructure and model deployment in production settings.

Standards

How we maintain course quality

Every cohort and every notebook goes through a review cycle before it reaches students.

Notebook testing

All notebooks are executed end-to-end on the current library versions before each cohort opens. Broken code does not reach students.

Peer review before publication

New content is reviewed by at least one other facilitator with relevant experience before it goes into the curriculum.

Post-cohort curriculum update

After each cohort closes, facilitator notes and student questions feed into the next version. No cohort receives unchanged material from more than a year prior.

Data privacy compliance

Student data is held in accordance with Thailand's PDPA. Enrollment data is used only for course administration and is not shared with third parties.

Cohort size limits

The Computer Vision cohort is capped at 30 students to keep feedback quality high. Once a cohort is full, the next intake opens rather than exceeding the cap.

Transparent pricing

All course prices are listed publicly in Thai Baht. There are no add-on fees after enrollment and no upsell requirements.

Expertise

Applied machine learning education in Bangkok

Synaptiq operates at the intersection of engineering practice and formal machine learning foundations. The courses focus specifically on PyTorch-based deep learning — not because other frameworks lack merit, but because depth in one tool is more useful than surface coverage of several.

Neural network development requires comfort with linear algebra, probability, and automatic differentiation. The Synaptiq curriculum introduces each of these at the point where they become necessary — not as prerequisites to clear before the interesting content begins, but as tools you pick up while building something real.

Computer vision is a productive first domain for applied deep learning. The problems are concrete, the evaluation metrics are interpretable, and the open model ecosystem is mature enough that a developer with moderate Python experience can reach non-trivial results within a single cohort. The Synaptiq Computer Vision course is structured around that arc.

Bangkok-based students can attend office sessions in person. Remote students participate in the same cohort structure through the online delivery format. All course material — notebooks, readings, evaluation exercises — is accessible to enrolled students regardless of location.