Three Stations. Three Distinct Skills.
Each Inferra programme has a defined scope, a project artefact as its endpoint, and tools you can use after the cohort ends. Choose the station that fits your schedule and current skill base.
← Back to HomeHow Each Programme Is Structured
Every Inferra programme is organised around a final project artefact rather than a topic list. From week one, participants know what they're building and why each part of the curriculum feeds into it. This isn't a stylistic preference — it changes how learning works in practice.
Content is delivered through a combination of written materials, code examples, and either live sessions (data engineering and notebooks) or one-on-one mentor calls (NLP). All programmes include an asynchronous channel where participants can raise questions between sessions.
Curriculum is reviewed on a fixed schedule twice per year. Updates reflect changes in the open-source Python and ML ecosystem — new library versions, deprecations, and shifts in how practitioners approach common problems.
Project-First Design
Scope is defined by the artefact you're building, not a module sequence.
Open-Source Toolchain
No proprietary platforms. Everything you build here runs in any standard Python environment.
Scheduled Updates
Curriculum reviewed twice a year against current ecosystem state.
Small Cohorts
Capped sizes preserve peer-review quality and mentor attention.
Data Engineering Foundations for ML
A ten-week practical course covering data ingestion, transformation, feature engineering, and pipeline-building practices using open-source tools. The course is project-driven, with each participant building a small end-to-end pipeline that loads, transforms, and serves features for a simple model. Includes weekly live discussion and a peer code-review channel.
- Weekly live discussion sessions with instructor
- Dedicated peer code-review channel per cohort
- End-to-end feature pipeline as final artefact
- Open-source tools: Python, Pandas, SQLAlchemy, Prefect
- Asynchronous question channel between sessions
Process Steps
- 01.Data source survey and ingestion planning
- 02.Raw data loading and schema analysis
- 03.Transformation and cleaning layer construction
- 04.Feature engineering and output serving
- 05.Pipeline orchestration and code review
Process Steps
- 01.Text preprocessing and tokenisation decisions
- 02.Embedding approaches — word vectors to transformers
- 03.First artefact: classification or tagging pipeline
- 04.Transformer fine-tuning with evaluation methodology
- 05.Second artefact: generation or search application
Natural Language Processing Mentorship
A four-month mentored programme on classical and modern natural-language techniques — text preprocessing, embeddings, transformer-based models, and basic evaluation methodologies. Participants meet a mentor fortnightly and build two project artefacts that they keep for their portfolio. Curriculum updates twice a year to reflect open-source ecosystem changes.
- Fortnightly one-on-one mentor sessions
- Two portfolio-ready NLP project artefacts
- Curriculum reviewed twice yearly
- Covers classical methods through modern transformers
- Async cohort channel between sessions
One-Month Crash Programme: Notebooks & Visualisations
A one-month intensive on building clean notebooks for analysis and presentation — layout, exploratory plots, narrative comments, and reproducibility practices. Suitable for participants with some Python experience who want to produce more readable analytical work. The programme finishes with each participant presenting a single notebook to the cohort.
- Requires basic Python experience
- Focus on layout, narrative, and reproducibility
- Cohort notebook presentation at programme close
- Compact four-week schedule, low time commitment
- Tools: Jupyter, Matplotlib, Seaborn, Plotly
Process Steps
- 01.Notebook structure principles and cell organisation
- 02.Exploratory analysis with narrative markdown
- 03.Plot construction and styling for readability
- 04.Reproducibility: seeds, environment files, comments
- 05.Cohort presentation of finished notebook
Which Programme Fits?
Use this matrix to compare what each station covers and who it's designed for.
| Feature | ST-01 Data Engineering | ST-02 NLP Mentorship | ST-03 Notebooks |
|---|---|---|---|
| Duration | 10 weeks | 4 months | 1 month |
| Price | ฿5,400 | ฿20,400 | ฿3,600 |
| One-on-one mentoring | — | — | |
| Live group sessions | Presentation only | ||
| Portfolio artefacts | 1 pipeline | 2 NLP artefacts | 1 notebook |
| Required Python level | Basic Python | Basic Python | Basic Python |
| Best for | Data/ML engineers | NLP specialists | Analysts & researchers |
Shared Across All Programmes
Privacy of Participant Data
Enrolment data and submitted work are not shared outside Inferra. Participant information is used only for programme management.
Open-Source Only Tooling
Every tool, library, and environment used in Inferra programmes is open-source. No subscriptions or licences are required from participants beyond what runs in a standard Python setup.
Defined Scope Per Cohort
Each cohort receives a written scope document at start. Participants know the artefact they'll produce, the sessions they'll attend, and the tools they'll use before the first week begins.
Honest Response to Questions
Pre-enrolment queries receive a direct answer. If a programme isn't suitable for a specific background or schedule, we say so rather than pushing an enrolment that won't serve the participant.
Twice-Yearly Curriculum Updates
All three programmes are reviewed on a fixed schedule. Outdated tool references, stale examples, and methodology sections that no longer reflect current practice are updated before the next cohort intake.
Capped Cohort Sizes
Inferra limits the number of participants per cohort to preserve meaningful peer interaction. When a cohort reaches capacity, the next intake is opened rather than expanding the current group.
Programme Fees
All fees listed in Thai Baht. Payment details provided after enrolment confirmation.
Notebooks & Visualisations
1-month programme
- Clean notebook construction
- Exploratory visualisations
- Reproducibility practices
- Cohort presentation
Data Engineering
10-week programme
- Pipeline design and construction
- Weekly live discussion
- Peer code-review channel
- End-to-end feature pipeline
NLP Mentorship
4-month programme
- Fortnightly 1-on-1 mentoring
- Two portfolio artefacts
- Transformers and embeddings
- Twice-yearly curriculum updates
Ask About Upcoming Cohorts
Send an enquiry with the programme you're considering and we'll respond with dates and schedule details within one working day.
Contact Us