Data Anonymisation & Privacy

A personal data inventory, pseudonymisation and masking that hold up, minimised pipelines, erasure that reaches the warehouse, small-number suppression and tested anonymisation, so analytics stays useful without exposing people.

Free · Mac (Apple silicon) and Windows · v1.0.5 · Other platforms

Then open the Marketplace in the app, search for Data Anonymisation & Privacy, and click Install. It goes into your own files, and works offline from then on.

What it installs.

1

Pillar

1

Area

50

Projects

197

Starter tasks

📊 Data & Analytics

Data Anonymisation & Privacy

Protecting personal data in analytics through minimisation, pseudonymisation, aggregation and access controls while keeping it useful.

  1. Personal data inventory of the warehouse
  2. Identifier taxonomy for direct, quasi and sensitive fields
  3. Analytics purpose register for personal data
  4. Data minimisation at the point of extraction
  5. Keyed pseudonymisation token design
  6. Privacy escalation route with the data protection officer
  7. Personal data out of development and test environments
  8. Privacy impact assessment for the analytics platform
  9. Shared definitions of anonymised and pseudonymised
  10. Monthly scan for new personal data columns
  11. Quarterly review of unmasked column access
  12. Masking policy coverage check
  13. Erasure requests carried through to analytics tables
  14. Pseudonymisation key custody and rotation
  15. Privacy review gate for new datasets and dashboards
  16. Query logs and BI extracts scrubbed of personal data
  17. Break-glass access to raw identifiable data
  18. Small number suppression rule for reports
  19. How linkage attacks re-identify people
  20. k-anonymity, l-diversity and t-closeness in practice
  21. Differential privacy fundamentals for analysts
  22. Why hashing an email is not anonymisation
  23. Synthetic data, what it can and cannot do
  24. Privacy-enhancing technologies landscape review
  25. Reading your regulator's anonymisation guidance
  26. Privacy training for the analytics team
  27. Choosing a tokenisation or data masking tool
  28. Aggregated reporting layer for business users
  29. Replacing email joins with a pseudonymous customer key
  30. Free-text redaction before loading
  31. Coarsening location data and IP addresses
  32. Date of birth and timestamp generalisation
  33. Motivated intruder test before calling data anonymous
  34. Analytics event payloads cleared of personal data
  35. Privacy audit and regulator inquiry readiness pack
  36. Releasing a de-identified dataset to a research partner
  37. Re-identification red team exercise
  38. Analytics privacy incident tabletop exercise
  39. Annual analytics privacy report to leadership
  40. Ethics approval and data plan for a study with personal data
  41. Statistical disclosure control for published tables
  42. Privacy by design checklist for a new pipeline
  43. Cross-border analytics and data residency map
  44. Bringing an acquired company's analytics under your privacy rules
  45. Special category data in a restricted analytics zone
  46. Children's data in product analytics
  47. Privacy budget ledger for a recurring private release
  48. Training data privacy for machine learning models
  49. Data clean room pilot with a measurement partner
  50. Automated re-identification risk scoring

Where it starts.

Each project arrives with its purpose, milestones and first tasks written out. Here is how the first three read.

Personal data inventory of the warehouse

This is a team plan, best installed into a shared workspace so everyone with a role in it can see and own their tasks. Most analytics platforms hold more personal data than anyone thinks: email addresses in event payloads, phone numbers in support exports, birth dates in a CRM sync nobody uses.

First tasks

  • Export the list of schemas, tables and columns from the warehouse information schema
  • Search column names for patterns such as email, phone, address, dob and ip
  • Sample 100 rows from each unclear free-text or JSON column
  • Tag each personal data column with its data type and source system
  • Ask the agent to summarise the inventory into a one-page view by source system

Identifier taxonomy for direct, quasi and sensitive fields

Removing names and emails does not make a dataset safe, because postcode, birth date and sex together can single out most people in a population. Sorting every personal column into direct identifiers, quasi-identifiers, sensitive attributes and non-identifying fields gives engineers a default rule for each class instead of a fresh debate for every column.

First tasks

  • Write one-line definitions of direct identifiers, quasi-identifiers and sensitive attributes
  • Assign each inventoried column to a class and flag the disputed ones
  • Agree a default treatment per class with the privacy lead
  • Apply the class as a tag or label on each column in the warehouse

Analytics purpose register for personal data

Privacy law in most places ties personal data to a stated purpose, and analytics teams are where purposes quietly stretch: data collected to deliver orders ends up feeding a churn model. A register of each analytical use, the fields it needs and the basis your privacy team has recorded for it lets you answer 'why do you have this' in one lookup, and catches new uses before they ship.

First tasks

  • List the dashboards, models and reports that use personal data
  • Write the fields and purpose for each use in one sentence
  • Send the register to the privacy team to record a basis for each use
  • Review the register for new and retired uses with the privacy team @recurring(yearly)

Make it yours.

Install Data Anonymisation & Privacy, then tell Aurum the date you're aiming for. It checks whether the plan fits the weeks you actually have.

Free · Mac (Apple silicon) and Windows · v1.0.5 · Other platforms

Blueprints are general information to help you organise your own work. They are not medical, financial, legal or other professional advice. See the Terms.