Flagship · 7 weeks
The Analytics Data Contract
For analytics teams who already ship events and need governance that survives the next reorg. You work on your own stream. We mark it in public, at the desk.
What you should be able to do afterwards
- Produce a one-page contract for one event family, including owners and downstream copies.
- Reject or rename properties that smuggle identifiers, including in the event title itself.
- Carry a consent flag through at least one warehouse hop without inventing a second truth.
- Show a retention clock that counsel in Thailand can read without a glossary.
- Run a change-request ritual that product, analytics, and warehouse can all sign.
Who it is for
Product analysts, analytics engineers, and analytics leads who can change tracking or warehouse models — or who sit next to someone who can. If you only consume dashboards, this path will feel like someone else’s job, because it is.
English materials; discussion may move between English and Thai. Remote seats are normal. Rayong in-person is optional and does not change tuition.
Modules
- Inventory before ideology. List events, owners, and copies. Discover the laptop CSV. Do not redesign the taxonomy yet.
- Naming as a control surface. Who may invent a property. How event titles leak PII. A written rule the squad can refuse to break.
- Consent flags that survive downstream. Flags computed after ETL are theatre. Design one that still means something in the warehouse.
- Retention clocks and warehouse copies. PDPA-aware windows, including debug extracts and partner feeds. Clocks, not slogans.
- Access, masking, and analyst roles. What “need to debug” is allowed to see. Row filters versus deleted columns.
- Change requests without breaking dashboards. Versioning names, deprecation windows, and who signs the request.
- Audit evidence a legal team can read. Assemble the week’s artefacts into a pack that does not require our presence to explain.
Instructor
Nalinee Prasert
Nalinee led product analytics at a regional commerce group before opening the Rayong desk. She still marks every cohort’s week-two owner map herself. She is not a practising lawyer; PDPA modules are taught as operational habits, with a standing invitation to bring your own counsel into office hour.
Informational tuition
฿48,500 per seat for the seven-week Rayong Cohort. A team desk of up to five, working the same stream, is typically ฿186,000. Field Notes (recordings and templates only) is ฿12,900. No payment is taken on this website. See pricing and the refund policy.
Questions we actually get
Do I need SQL fluency?
Enough to read a model and name a column. We do not teach SQL from zero. If warehouse lineage is your only gap, the four-week lab is a better fit than this flagship.
Will this certify us for GDPR?
No. That is a real limitation, not a modest disclaimer. We teach PDPA-aware retention and consent flags that travel with events. We do not certify GDPR programmes, issue legal opinions, or act as your Data Protection Officer. If your primary user base is in the EU, treat this as complementary training for the analytics craft — not a substitute for counsel in that jurisdiction.
Can we bring a secret stream?
Yes, with a written note that the room will see event names. We never publish student schemas. If your counsel forbids any external eyes, book a Resident Desk under NDA instead of a public cohort.
What if I miss a week?
Recordings of the lecture portion exist. Desk review does not. Missing more than one live desk usually means you should defer to the next cohort rather than collect a thin certificate.
From people who sat this path
Week five’s change-request ritual is now how we rename anything. I still think the masking studio should be a full extra day; we ran out of time on Looker row filters.
★★★★☆
Verified seat · January 2026. “Owner map is on our wall. Consent lab was the surprise. Lineage week assumed more dbt than we had.”