Skip to content
View AKARandy's full-sized avatar

Block or report AKARandy

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Content in all repositories owned by your account will be closed.
Maximum 250 characters. Please don’t include any personal information such as legal names or email addresses. Markdown is supported. This note will only be visible to you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
AKARandy/README.md

Hi, I'm Alief (AKARandy)

Mathematics graduate from Institut Teknologi Sepuluh Nopember (ITS), class of 2026. Most of my work sits in the unglamorous parts of machine learning: cleaning messy data, scraping sources that have no API, and building pipelines that catch their own drift and retrain before anyone has to notice. On quieter days I build small games in Godot.

If you want to talk data, scraping or MLOps, reach me at aliefrandiansyahp@gmail.com.


Featured work

Project What it does
QUACKO QA suite for an OCR receipt reader: Flask plus Tesseract service covered by pytest and Playwright suites (API contract, degradation, regression, UI), CI on every PR and push, reports on an evidence site
Lights-vs-Numbers Data analysis, Indonesia GDP vs independent signals: official BPS GDP (42-quarter sourced dataset) against PortWatch shipping, search trends and night lights. Annotated Plotly charts, Streamlit dashboard plus static mirror (live demo), CI green
Self-Retraining-Model-Platform MLOps loop where drift detection triggers automatic retraining: conditional promotion, zero-redeploy serving. Argo, MLflow, DVC, Optuna on Kubernetes. Test case: house prices
tanya-sekolah "Ask the school": LLM Q&A plus tuition (SPP) prediction over Indonesian school data
Qwen-on-Kaggle-TPU-via-zrok-and-API Serving Qwen 27B (BF-16) on free Kaggle TPUs through zrok as an OpenAI-compatible API
kacasaham Weekly Telegram digest scoring IDX stocks on trust and resilience, evidence link per claim. Sectors Hackathon 2026
DAWG-ID Macro shock vulnerability index for 38 Indonesian provinces, choropleth portal plus dataset. Top 20, Microsoft AI Impact Challenge Datathon
InternProject-SPP_Sekolah Predicting school tuition fees from open-source data
Education data scrapers Crawlers for Indonesian open data: Dapodikdasmen, PDDIKTI, Sekolah Kita, Verval Yayasan, LinkedIn jobs

Toolbox

Languages: Python · Java · GDScript (Godot) · LaTeX

ML / Data: Machine Learning · Time Series Forecasting · Data Analysis · Web Scraping · LLMOps · OCR · pandas · Plotly · Streamlit

Infra: Docker · Kubernetes · Argo Workflows · MLflow · DVC · Optuna · GitHub Actions

Certifications

  • Mathematical Data Analytics Bootcamp (Matematika ITS, Sep 2026)
  • Getting Started with Docker (Simplilearn, Sep 2026)
  • Python for Data Science and Machine Learning Bootcamp (Udemy, Nov 2024)
  • Intro to Machine Learning (Kaggle, May 2023)
  • Scientific Computing with Python (freeCodeCamp, Jul 2021)

GitHub stats

Top languages

GitHub streak

Connect

LinkedIn Instagram Email


Pinned Loading

  1. Self-Retraining-Model-Platform Self-Retraining-Model-Platform Public

    Self-retraining model platform: drift detection triggers automatic retraining, conditional promotion, zero-redeploy serving. MLOps on Kubernetes (Argo, MLflow, DVC, Optuna). Test case: house prices.

    Python

  2. tanya-sekolah tanya-sekolah Public

    Sekolah Assistant: LLM Q&A + SPP prediction over Indonesian school data

    Python

  3. DAWG-ID DAWG-ID Public

    HTML 2

  4. Qwen-on-Kaggle-TPU-via-zrok-and-API Qwen-on-Kaggle-TPU-via-zrok-and-API Public

    Qwen 3.8 27B BF-16 on Kaggle via zrok and OpenAI-Compatible API

    Jupyter Notebook 1

  5. Lights-vs-Numbers Lights-vs-Numbers Public

    Data analysis: Indonesia official GDP vs independent activity signals (shipping, search trends, night lights). Python, pandas, Plotly, Streamlit.

    Python

  6. QUACKO QUACKO Public

    QUACKO: multi-layer QA suite for a real Flask+Tesseract OCR SUT (Committed 20 real receipts, API+model+UI gates, mkdocs reports site)

    Python