ZIDAN MUBARAKBANDA ACEH, ID
Loading portfolio
Loading
ZIDAN MUBARAK
ENID
All work

02 / 06

Machine learning

Pacu Pintar

An AI learning-analytics platform where a Random Forest labelled study pace and a FastAPI service turned it into study advice.

Live (opens in a new tab)Source (opens in a new tab)
Role
ML engineer, two-person ML team
Year
2026
Stack
Python · scikit-learn · Random Forest · FastAPI · LLM · Next.js · Hapi.js
Reading time
Project note
Team
Asah led by Dicoding capstone; ML team of 2 inside a full-stack team
The public Pacu Pintar landing page with the Indonesian headline for its structured coding learning paths.
Fig. 01 · Live captureThe public Pacu Pintar landing page. The signed-in student dashboard is not captured.

01 · Overview

What it is

An AI learning-analytics platform where a Random Forest labelled study pace and a FastAPI service turned it into study advice.

The ML half of a web platform for learning analytics. A classifier sorts students into three study-pace categories; a FastAPI service exposes the model and asks an LLM for a personal study plan. It won Best Capstone Project in the cohort.

02 · Highlights

Highlights

  1. 01

    Random Forest classifier for three study-pace categories (consistent, fast, reflective) at ≈90% accuracy.

  2. 02

    FastAPI service serving the classifier plus LLM-generated personal study suggestions.

  3. 03

    Integrated with the team's Next.js and Hapi.js web application.

  4. 04

    Best Capstone Project, Asah led by Dicoding.

03 · Stack

Stack

  • Python
  • scikit-learn
  • Random Forest
  • FastAPI
  • LLM
  • Next.js
  • Hapi.js

Sources