01 — Case Study
Wikalpa AI Assessment & LMS
Client
Wikalpa
Category
EdTech & AI Solution
What we built
NexLMS Core deployment with automated essay scoring and AI conversational practice agents
Stack
NexLMS Core (React + Vite frontend, FastAPI backend, MongoDB, AI scoring pipeline)
What made this challenging
Wikalpa needed an AI assessment system that could grade open-ended essay responses in Bahasa Indonesia — a language with complex morphology that most off-the-shelf NLP tools handle poorly. We built a custom scoring pipeline using fine-tuned BERT models on Indonesian academic text, achieving reliable automated grading that correlates with human assessors.
Outcome
The platform is live and serving Wikalpa's students with AI-powered essay scoring and conversational practice agents. Students receive immediate feedback on their writing, and the system identifies knowledge gaps for targeted review.