Fresh graduate in Informatics Engineering with a strong interest in Data Science, software development, and data processing. Experienced in IT support and data-related projects, with a continuous drive to learn and develop new skills.
I'm Elvina, an Informatics Engineering graduate from UHAMKA. I work where data gets messy: cleaning and wrangling datasets, exploring them, and building machine learning models in Python that turn them into predictions.
I trained as a data scientist through the MSIB program at BISA AI Academy, kept the computer labs at UHAMKA running as a laboratory assistant, and handled correspondence and finances for Developer Student Club UHAMKA. I like work that needs both care and analysis, and I learn fast when a problem is new.
PT. BISA AI Academy · MSIB
Universitas Muhammadiyah Prof. Dr. Hamka
Developer Student Club UHAMKA
Jungle Leiter
Airline passenger satisfaction app
Passengers fill in a flight survey and a machine learning model predicts whether they were satisfied; admins follow the results on a dashboard. Built by a team of six as the capstone project of the MSIB program at BISA AI Academy.
Indonesian sign language to text & speech, in real time
My undergraduate thesis: a webcam app that recognises 15 BISINDO (Indonesian Sign Language) words as they are signed and reads them aloud in Indonesian. MediaPipe Holistic tracks body and hand landmarks, and a hybrid Transformer Encoder + LSTM model trained with CTC classifies the motion sequence.
Product review prediction with NLP
Classifies product reviews as positive or negative using natural language processing, so a company can read customer perception at scale and use it to improve product quality and plan marketing. Built during the Data Science track of MSIB at BISA AI Academy and published on the BISA AI portfolio.
Finding objects in a photo with template matching
An image processing notebook in OpenCV. Its template matching step slides a small template
across a photo with cv2.matchTemplate, takes the strongest correlation and
boxes the match. Built for the Image Processing course of MSIB at BISA AI Academy and
published on the BISA AI portfolio.
Three-class review sentiment, served through a Flask API
A sentiment model that sorts product reviews into positive, neutral or negative, taken past the notebook: the trained pipeline is saved with joblib and served by a Flask endpoint that takes a review and returns its sentiment. Built for the Data Science course of MSIB at BISA AI Academy.
/predict endpoint, built in PyCharmRegression, classification and clustering models in Python, evaluated and tuned on real datasets. Used for prediction, sentiment analysis and computer vision projects at BISA AI.
My main language for data processing, analysis and model code, written in notebooks and scripts.
Cleaning, reshaping and preprocessing raw data until it's ready to analyse.
Fast, accurate input and record keeping.
Installing and maintaining software, keeping networks up, and diagnosing hardware, software and connectivity problems for users.
UI/UX design for websites in Figma, plus posters, presentation decks and other visual materials in Canva.
Informatics Engineering. Coursework across programming, data processing, artificial intelligence and software development.