EdTech / FMCG
EdTech
A major international consumer brand relied on external sales and service professionals who needed strong product knowledge but were hard to reach through a corporate learning system. Every campaign wanted its own training, and every one of them was turning into its own application.
What was built
The pieces, and how they fit
All of it is configuration on a single instance.
Course content and product information in collections, teaching approach and campaign rules in skills pulled into context when relevant
Learning agents that read only the collections and skills granted to their programme, so two campaigns on one instance cannot see each other's material
A store with a schema holding one entry per participant, so a session resumes rather than restarts
Functions for game logic, scoring, invitations and notifications
Named queries behind the reporting interface, giving real-time participation data without direct database access
A namespace per programme on the same deployment, so a new learning application is a configuration change
The result
New programmes are assembled from reusable parts instead of built from scratch. In the original implementation a new game could be configured in under 30 minutes.
AI agents
Skills
Collections
Templates
State stores with schema
Python functions
Named queries
Namespaces per programme
Get started