Solution Design - MVP

Scenario:

NISHTA is running a course - it may have up to 1 million users enrolling to the course.

Assumption: We will have a course that is being taken up by 500 - 1000 users…

Category - Course mapping: one category per batch

Category creation:

Categories:
Name of the category: <tenant_id>-<course_id>-<batch_id> - will this kind of name be OK?
- Option 1: use names instead of identifiers
  - should we use <tenant_name>-<course_name>-<batch_name>
  - NCERT - Early Childhood Education - Batch 01
- Option 2: hierarchical category structure
  - Tenant Name > Course Name > Batch Name
This will be a root category
Visible only for users enrolled to the batch
the association between batch and category will be stored on the course object in Sunbird
- Batch:b1 - Category:11
- Textbook:t1 - Category:22
Whitelisted Tags:

Moderators for the category:

Batch mentors will get the moderators access for the category
if 10000 (1% of 1 million) users are posting questions in the forum, how will tenant manage the moderation?

User Mapping:

Sunbird User Profile: identifier (uuid), username (alphanumeric), email or phone, first name and last name
NodeBB: identifier, username and email (username@domain)
Create user in NodeBB when created in Sunbird (scale, not for MVP)
- listen to create user event and create user in NodeBB

Category access to course batch members:

One time batch job for existing
Streaming:
- listen to the user enrolment event
user is already created in NodeBB
- use the username or identifier (whatever nodeBB API supports) to add privilege
if user is not yet created in NodeBB (for MVP)
- create user
- and add privilege to the user

Use Tags for moderation & answers:

tag-a and tab-b
- instead of directly tagging the topic/post, we can infer the tags from user profile
topic is flagged
- highlight to user based on tags & weightage
when a new question is posted
- show it to the users based on tags & weightage

Tags:

Accountability:

use telemetry to determine KRA
- metrics - percentage of posts replied within a day/hour/week, etc and percentage of flags addressed within a day/hour/week, etc
ensuring replies within an SLA
ensuring flags are addressed with an SLA

Closing a category:

Category Types:

Closed (E.g. Course batch)
- Unique group per category
Open (E.g. Book content)
- Universal access based on registered-users group
Restricted (E.g. one state allowing users from another states to explicitly join their category and interact)
- Universal access based on tenant-registered-users group

Use cases:

Discussions