Three Programmes.
One Clear Path into AI Development.
Each programme at Tensoria is built around deliberate practice — structured modules, realistic exercises, and thoughtful feedback at every stage.
Back to HomeA Studio Approach to Learning
Rather than passive content delivery, Tensoria organises its programmes around hands-on studio work. Each topic is introduced through a brief conceptual framing, then practised through structured coding exercises that grow in scope as the programme progresses.
Concept
Each module begins with a focused introduction — just enough theory to make the practice meaningful.
Practice
Coding exercises are staged progressively — earlier tasks build the foundation for more involved later work.
Review
Work is reviewed with specific notes — not scores, but observations about approach and reasoning.
Synthesis
Each programme closes with an integrative project that ties earlier work into a coherent, documented output.
Deep Learning & Neural Networks
A structured course on neural network design and training. The programme moves through foundational architectures — perceptrons, convolutional networks, recurrent models — with progressively more involved coding exercises at each stage. By the final modules, participants are designing and training multi-layer networks from a defined specification.
What the programme covers:
- Feedforward networks, activation functions, and gradient descent
- Convolutional architectures for image-based tasks
- Sequence modelling with recurrent and transformer-based networks
- Regularisation, optimisation strategies, and training diagnostics
- Final practical: design and train a neural network to a written brief
฿5,100
Generative AI Project Studio
A project-based studio where participants design, build, and document a small generative application from start to finish. Rather than covering generative AI as a broad topic, this programme organises learning around a single sustained project — every module contributes to a concrete, deployable output.
Studio project scope:
- Project scoping, architecture planning, and tool selection
- Prompt engineering, fine-tuning fundamentals, and retrieval-augmented patterns
- Iterative build cycles with structured feedback between rounds
- Evaluation methods and quality review for generative outputs
- Project documentation: design decisions, limitations, and future directions
฿19,000
MLOps & Deployment
A course on packaging, serving, and monitoring models — taught through realistic deployment exercises. Each module simulates a stage in a production pipeline, from containerising a trained model to setting up monitoring dashboards and handling model drift.
Deployment topics covered:
- Model packaging: environments, dependencies, and reproducibility
- Serving patterns: REST APIs, batch inference, and async pipelines
- CI/CD practices for ML — version control, testing, and pipeline automation
- Monitoring strategies: performance logging, drift detection, and alerting
- Infrastructure cost management and scaling considerations
฿34,000
Choosing a Programme
Each programme is self-contained, though they build naturally on each other. Here's a quick comparison to help you identify the right starting point.
| Feature | Deep Learning ฿5,100 |
Generative AI Studio ฿19,000 |
MLOps & Deployment ฿34,000 |
|---|---|---|---|
| Format | Structured course | Project studio | Exercise-based course |
| Primary focus | Neural network design | Building a gen. AI app | Production deployment |
| Coding exercises | |||
| Sustained project | — | — | |
| Deployment focus | — | Partial | |
| Project documentation | — | — | |
| Good starting point if… | New to neural networks | Ready to build something real | Need production-ready skills |
Start with Deep Learning if…
…you are building a foundation in neural networks and want a well-paced, structured entry point before exploring generative or deployment topics.
Start with Generative AI Studio if…
…you already have some ML background and want to produce a complete, documented generative project you can refer to in future work.
Start with MLOps if…
…you can already train models and want to strengthen your ability to package, ship, and maintain them in realistic deployment settings.
Programme Standards
A few principles that guide how every Tensoria programme is designed and delivered.
Data Privacy
Participant work and contact details are handled in accordance with Thailand's PDPA. Nothing is shared with third parties without consent.
Paced Difficulty
Exercise complexity increases incrementally. Early tasks are designed to be accessible; later ones require applying earlier learning in unfamiliar contexts.
Specific Feedback
Feedback is written per submission — not automated. Notes address how an exercise was approached, where reasoning was sound, and what to reconsider.
Ethics in Practice
All programmes include discussion of responsible AI use — bias, transparency, and appropriate deployment context. These are woven into exercises, not treated as a side topic.
Programme Pricing
All fees are in Thai Baht. Reach out to confirm availability and discuss any questions before enrolling.
Deep Learning & Neural Networks
Structured course from fundamentals to multi-layer network training.
- Modular structure with checkpoints
- Progressive coding exercises
- Written feedback per submission
- Final practical project
Generative AI Project Studio
End-to-end project: design, build, and document a generative AI application.
- Single sustained build project
- Iterative build cycles
- Structured review between rounds
- Full project documentation
- Ethics and evaluation module
MLOps & Deployment
Realistic exercises covering the full model-to-production pipeline.
- Packaging and containerisation
- Serving patterns and APIs
- CI/CD for ML pipelines
- Monitoring and drift detection
Questions before enrolling?
If you are unsure which programme fits your current level or goals, we are happy to talk it through. Send a message or reach us by phone.