Hi everyone, my name is Dila Ayu Prastita.
I am an Informatics Engineering student at the Sumatra Institute of Technology with a deep interest in Data Science and Data Analytics. I have experience as a Management Information System (MIS) student at PT Lampung Berkah, where I was responsible for analyzing company data, building interactive dashboards using Tableau, and developing business reports to support decision-making. My expertise includes data processing and visualization using Excel, SQL, and Python.
My interest in Machine Learning was realized through my active participation in various academic activities, including national seminars. I successfully won an award for best paper author, demonstrating my ability to process data, build analytical models, and communicate results systematically and concisely. This experience strengthened my understanding of the implementation of machine learning in real-world contexts, particularly as a strategic decision-making tool.
In the UI/UX field, I participated in the GEMASTIK 2024 competition, where my team and I redesigned the Customs and Excise application using a user research-based approach. We leveraged technologies such as AI and machine learning to improve the user experience. In addition, I am also active as a Teaching Assistant for the Human-Computer Interaction course, guiding students in understanding the principles of effective and user-friendly interface design.
I analyzed the company's operational and financial data using Microsoft Excel and SQL to identify patterns, trends, and potential efficiencies. I then visualized the results of this analysis through an interactive dashboard built using Tableau. This dashboard displays various Key Performance Indicators (KPIs), business performance trends, and other data insights in real time, supporting faster, data-driven decision-making.
Note : All numerical data in dashboards have been blurred and charts have been manipulated to maintain data confidentiality.
I developed a predictive model to estimate long-term business ceilings using a machine learning approach. The model was built in Python, utilizing the XGBoost algorithm and libraries such as NumPy, Matplotlib, and Scikit-learn for data cleaning, model training, performance evaluation, and result visualization. This model achieved a prediction accuracy of 81%, demonstrating its reliable performance in providing long-term financial projection estimates. The results of these predictions are used to support data-driven strategic business decision-making.
Note : All numerical data in dashboards have been blurred and charts have been manipulated to maintain data confidentiality.
I designed and implemented a rule-based chatbot on my company's website for automated customer support. The chatbot greets users, collects their name and WhatsApp number, which are stored in a database, and displays FAQs with automated answers. Questions outside the FAQ are directed to customer service via WhatsApp or email. This project integrates logic, databases, and web technologies to improve customer service efficiency.
My team and I developed a vegetable detection model for types such as potatoes, carrots, capsicum, and broccoli using digital image processing techniques (resize, normalize, augmentation, enhance contrast, grayscale) and deep learning methods with CNN architectures (CONV2D, Xception, VGG16). The dataset consisted of 400 images sourced from Kaggle, ensuring the model's ability to accurately recognize and classify vegetable types.
I created a professional online portfolio using HTML, CSS, and JavaScript, showcasing my biodata and experiences. The design emphasizes aesthetics and functionality, highlighting my skills in crafting user-friendly and visually appealing web solutions.
My team and I designed Nuju Coffee’s website for online ordering, utilizing Figma for UI/UX planning. We responded to customer feedback through surveys, conducted UX testing with SUS and UEQ, and presented the prototype to lecturers, highlighting our skills in Figma and user-centered design.
My team and I participated in the GEMASTIK 2024 competition, where we redesigned the Customs application and named it "CukaiPro" using Figma, incorporating advanced technologies such as AI and machine learning.
LIHAT YOUTUBE
My team and I developed a Telegram chatbot using GEMINI APIs, integrated with the company database to deliver responsive, human-like interactions. We analyzed requirements, designed solutions, and used Python to process database data. Testing ensured accurate, fast responses, with SQL logic providing actionable insights, and seamless backend connectivity supporting efficient communication.
My team and I developed a BBC news classification model, applying Natural Language Processing (NLP) techniques to compare Recurrent Neural Network (RNN) and Naive Bayes Classifier for category prediction. Using TF-IDF for text representation and SVD for dimensionality reduction, we analyzed sequential relationships (RNN) and probabilistic patterns (Naive Bayes). This project highlights the effectiveness of NLP-based algorithms for automatic text classification, specifically on the BBC news dataset.
My team and I developed an expert system model for disease diagnosis that utilizes medical ontology and the Random Forest algorithm. This system is designed to improve disease diagnosis accuracy by integrating structured medical knowledge through ontology and machine learning algorithms.
My team and I conducted sentiment analysis on the Pocket ITERA app, a student platform at the Sumatra Institute of Technology (ITERA) offering services like attendance, helpdesk, and e-counseling. Using NLP and Naive Bayes, I analyzed Google Playstore reviews to classify feedback into positive and negative sentiments, highlighting the app's strengths and areas for improvement. The results were visualized to provide actionable insights, helping developers enhance the app's quality and user experience.