Akmal Muzakki Bakir

QUALIFICATION

My Experience

Data Mining Laboratory Assistant
- Digital Talent Center Laboratory (November 2024 - Present)

Active as PIC of the Adikara 2024 Data Mining division
Active in opening the Adikara 2024 Data Mining competition section
Actively host the preliminary round of the Kaggle-based data mining competition
Active in managing the final round and championship announcements

Researcher
- Badan Riset dan Inovasi Nasional (July 2024 - Present)

Data Science Research

Developing a web-based application for phytoplankton segmentation and classification
Developing a deep learning model using various pretraining techniques
Utilizing datasets obtained from Belitung waters
Research theme: "Modeling Phytoplankton Segmentation and Classification in Belitung Waters"

Research Assistant

Active in paper research entitled "Modeling of Phytoplankton Segmentation and Classification in Belitung Waters" to be published in the Journal of Marine Systems
Conduct research to modify the YOLOv8 architecture
Modify the YOLOv8 architecture by replacing a number of backbones with the proposed backbone (such as the KAN method from the convolution layer, and others)
Conduct data augmentation and image processing using the enlighten method

Cloud Computing Engineer
- Bangkit Academy (September 2024 - December 2024)

Utilize Google Cloud services to support application need such as Firestore and Cloud Storage for Database Service
Build cloud architecture to illustrate all necessary components and technologies required by the applications and machine learning models.
Utilize Google Cloud services for the machine learning workflow, such as deployment machine learning model and Gemini Autocorrect using FastAPI for API development, Artifact Registry for build a Docker Image, and Cloud Run Container for CI/CD API Production
Utilize Google Cloud Service for the building a serverless infrastructure such as create quiz questions menu and produce tutorials videos menu to storage which are then processed for API production using Express JS, and App Engine for CI/CD API Productions
Build authentication services such as sign up and sign in based on cloud firestore and update profile photos based on cloud storage

Graph Database Research
- Big Data Laboratory Telkom University (August 2023 - September 2024)

Develop methods and tools for visualizing graph queries and their results, including techniques like graph layouts, node-link diagrams, and matrix representations.
Study approaches to track and record the origin and transformation of graph data, utilizing provenance graphs, annotations, and timestamps.
Develop techniques and systems for discovering relevant graph data from heterogeneous sources, incorporating strategies such as schema matching, entity resolution, and query expansion.

Practicum Assistant
- IF Laboratory Telkom University (September 2023 - June 2024)

Practicum Assistant of Object Oriented Programming

Assist students in creating classes and objects, including defining attributes and methods.
Guide students in applying key OOP concepts such as inheritance, polymorphism, abstraction, and encapsulation.
Provide suggestions and input to help improve the students' program code.
Evaluate the results of the OOP practicum by assessing the students' program code.
Provide feedback on the students' work and assign grades according to the assessment criteria set by laboratory assistants and lecturers.

Practicum Assistant of Programming Algorithm

Assist students in introduction to java programming language, variable and operators, and basic sintax of java.
Guide students in applying array, searching, and sorting concepts
Provide suggestions and input to help improve the students' program code.
Evaluate the results of the Programming Algorithm practicum by assessing the students' program code.
Provide feedback on the students' work and assign grades according to the assessment criteria set by laboratory assistants and lecturers.

Computer Vision Research
- AI Laboratory Telkom University (September 2023 - June 2024)

Explore the use of deep learning methods for various computer vision tasks, including multiple object tracking, video object segmentation, pedestrian trajectory prediction, visual localization, change detection, image retrieval, video forensics, and video anonymization.
Conduct research related to computer vision with a focus on object detection. The research title is "Trash Detection in Surface Waters using YOLOv8."

Teaching Assistant
- Telkom University (March 2023 - January 2024)

Teaching Assistant of Algorithm Complexity Analysis

Assist lecturers in teaching how to measure the efficiency of algorithms based on execution time and memory usage.
Use notations such as O, Ω, Θ, o, and ω to determine the upper bound, lower bound, and tight bound of algorithm complexity.
Create exercises and exams related to algorithm complexity.
Evaluate student learning outcomes on these topics.
Provide guidance and consultation to students who have difficulties in analyzing algorithm complexity.

Teaching Assistant of Programming Algorithm

Provide additional material using the Go language and pseudocode.
Deepen the course material through responses and interactions.
Assist students with their assignments.
Supervise exams and provide assessments for large assignments.
Guide students in completing their assignments.

Teaching Assistant of Discreate Math

Assisted lecturers in delivering material on logic, sets, relations, functions, combinatorics, and graphs.
Created exercises and exams related to these topics.
Evaluated student learning outcomes.
Provided guidance and consultation to students who had difficulty understanding discrete math concepts.

Tentor
- HIMA IF Telkom University (September 2023 - November 2024)

Tentor of Statistics and Probability

Teach basic concepts of probability theory, including sample space, events, random variables, chance functions, distribution functions, expectations, and stochastic processes.
Explain and provide examples of the applications of probability theory in various fields, such as statistics, distribution theory, computer science, and others.

Tentor of Algorithm Complexity Analysis

Teach how to measure and compare the efficiency and performance of different algorithms.
Cover topics such as asymptotic notation, time complexity, space complexity, auxiliary space, worst-case analysis, average-case analysis, best-case analysis, and various types of complexity.

Data Engineer
- Kalbe Nutritionals (August 2023 - September 2023)

Implement the basic concepts and important techniques in data warehousing and data analytics, especially for beginners.
Apply these concepts and techniques in the realm of data warehousing competency.
Implement Extract, Transform, Load (ETL) processes using tools such as Pentaho and Airflow.
Create database connections, run ETL procedures, and integrate Pentaho with Google Cloud Platform (GCP) for effective data warehousing implementation.

My Project

🔗 Predict Salary App

This machine learning website predicts salary based on factors such as country, lecturer status, employment status, occupation, gender, and experience.
It is built using sequential deep learning and traditional machine learning models, including Gradient Boosting.

🔗 TMDB Movie Recommendation System App

This deep learning-based movie recommendation system uses sentiment analysis.
Built with Python and JavaScript frameworks, it uses AJAX and jQuery for dynamic interaction, and HTML, CSS, and Bootstrap for the frontend.
Film data is collected via web scraping from IMDb.
Sentiment analysis is performed using a Bi-LSTM sequential deep learning model, and the system includes an autocomplete feature for user convenience.

🔗 ROOM

The UI/UX design for a temporary accommodation search application (boarding houses, apartments, villas, hotels) was created using Figma.
The design focuses on providing an intuitive and engaging user experience for easier search and selection of temporary housing.

🔗 TANGGUH: Multi Models Machine Learning Application accompanied by Support Assistant

This machine learning website predicts stunting and provides nutritional advice through chatbots.
It uses generative AI with Gemini and GPT-4 language models.
The application predicts stunting based on entered parameters and offers suggestions from the generative AI model.

🔗 Telco Customer Churn Dashboard

This customer churn analysis employs a gradient boosting approach and exploratory data analysis.
The goal is to identify potential churners based on entered parameters and conduct in-depth analysis through feature extraction and extensive data exploration.

🔗 Trash Detection in Surface Waters using YOLOv8

Aquatic environments are polluted by trash like bottles and plastics, which threatens marine habitats.
Coastal nations contribute significantly, with 4.8 to 12.7 tons of trash entering the ocean annually.
YOLOv8 architecture was used for single-class trash detection in surface water.
Evaluations using the FloW-Img and WaterTrash datasets showed that the YOLOv8x model achieved the highest mAP50 score of 0.923, with minimal performance variation across different model sizes.

🔗 Network Structure Analysis of #FreePalestine Communication on Social Media X Using Social Network Analysis

This analysis aims to assess the communication, information dissemination, and media influence in Indonesia regarding the Palestine-Israel conflict on platform X.
Qualitative research methods were used, including Social Network Analysis (SNA) of international online media.
Python’s NetworkX library was utilized to visualize graphs and interactions, and to analyze centrality measures such as degree, betweenness, and closeness centrality related to the conflict.

🔗 Community Structure Analysis K-Pop Fans on Social Media X Use Social Network Analysis Girvan-Newman Algorithm

This study analyzes the community structure of K-Pop fans on platform X using Social Network Analysis (SNA) with the Girvan-Newman Algorithm.
Data from tweets hashtagged #kpop were collected between November 28, 2023, and December 6, 2023.
After preprocessing and adding "total_interaction," a network graph was created with NetworkX.
The Girvan-Newman algorithm identified 12 communities with a peak modularity of 0.79268.
Entities like MnetMAMA, enhypen_members, and bts_twt emerged as key community hubs, highlighting their role in fan engagement and cohesion.

🔗 Sirekap Optical Character Recognition using RCNN Method

The image-based text recognition system for the Sirekap case uses Optical Character Recognition (OCR).
It leverages OpenCV with Python and implements CTC loss + RCNN in the tf-keras module.
The system detects text in images, resulting from sound calculations.

🔗 Satellite Image based Area Damage Analysis using BiT-M and ViT-M

This research uses Big Transfer Medium (BiT-M) and Vision Transformers (ViT-S16) models for classifying damaged areas in satellite images.
The BiT-M R101x3 model achieved the highest classification accuracy of 83.22%.
Transfer learning with pre-trained models (BiT-M R50x1, R50x3, R101x1, R101x3) proved effective.
The study uses satellite image datasets from before and after disasters to classify areas with and without damage.

🔗 Pemilu Multiclass Text Classification using NNLM and Variational BERT

This research focuses on multiclass text classification using transformer architectures (BERT, RoBERTa, IndoBERT, IndoBERTweet) and deep learning models (pretrained NNLM, Bi-LSTM).
It also employs ensemble gradient boosting methods from traditional machine learning.
Preprocessing techniques include word importance, back translation, synonym replacement, and feature extraction.

🔗 MOODIFY: Anxiety and Stress Detection System Deep Learning Based with an Assistant Supporters

This research develops a web-based application using AI to help users manage stress and anxiety.
It employs AI models such as Ensemble Gradient Boosting, Stacked Bi-LSTM, Gemini, Gemma, Llama2, GPT-4, and RAG to predict stress levels and provide recommendations.
The application uses synthetic tabular data and relevant text from Kaggle, which are preprocessed and used to train the AI models.

🔗 Campaign Topic Modeling and Social Networks on Social Media X

This research focuses on the analysis of election topics from large datasets spread across social media X with a data collection period of 1 month
This research uses 2 models, namely topic modeling (dynamic topic modeling) and social network analysis (community detection)
The topic modeling architecture used is IndoBERTopic (using embedding from IndoBERTweet)

🔗 Phytoplankton Segmentation and Classification Modeling in Belitung Waters

This study focuses on the automatic identification of phytoplankton species using two modeling flows: segmentation followed by classification.
The segmentation models used are DeepLabV3+, U-Net, and SegNet with a ResNet50 encoder, while the classification models include Vision Transformers, BigTransfer, Swin Transformers, RegNet, and ConvNext with a concatenate layer on the transformers architecture.
The system workflow progresses from raw image input to segmentation, contour detection, cropping, and classification.
Segmentation achieves a top IoU score of 93%, and classification reaches 100% accuracy.
A web-based application was also developed using HTML and CSS for the frontend, and JQuery and Flask for the backend.

🔗 Signify - Sign Language Learning & Communication Aid

This application is mobile-based integrated with AI technology.
This application works as a real-time hand gesture detection accompanied by AI-based autocorrect. This application also includes video tutorials and quizzes for learning to develop.
The technology used is Android Jetpack to build the frontend system. Fast Api, Express.js, and several products from Google Cloud Platform to build the backend system, and CNN and Mediapipe technology to build the AI system.

MY SKILLS

I have been made several skills. Here are some of my selected skills

PROGRAMMING

Slide your mouse here

Python
Java
Go
Javascript
HTML/CSS
.
.
.

PROGRAMMING

Slide your mouse here

Until now, I am used to using Python programming language because I dream of becoming a data scientist or at least a career in data, but in some cases, I also use javascript programming languages (sometimes, I also used Javascript library like react or next for the simple project), HTML, and CSS to design website interfaces as a place for me to put my portfolio, I have also used go and javascript languages as rest APIs or backend tests, In addition, I use Java to design mobile apps (basics) to learn new material in android studio.

DATABASES

Slide your mouse here

MySQL
PostgreSQL
MongoDB
Neo4j
Nebula Graph
.
.
.

DATABASES

Slide your mouse here

I can implement SQL languages including DDL and DML. The database I'm used to is MySQL, but in some cases, I also use PostgreSQL because of its language structure that is almost close to SQL language and many more accessible features than MySQL. I've also learned MongoDB, but it's still a basic level.

DESIGN

Slide your mouse here

Figma
Adobe After Effect
Filmora
Canva
.
.
.
.

DESIGN

Slide your mouse here

In Figma, I once made a UI of a rented residence application, In Adobe After Effect, about a few years ago, I made a creative opening video together with Filmora. For small design projects, I usually use Canva. a few years ago, I created a blog on WordPress with the theme of travel about life.

BASIC SKILLS

Slide your mouse here

Data Science
Public Speaking
Networking
Critical Thinking
Leadership
Problem Solving
.
.

BASIC SKILLS

Slide your mouse here

I have several portfolios that I store on Github, Among some of these portfolios, most of them portfolios that I keep are related to data, which can be known because the domain file is Jupyter, some of these projects are based on machine learning, data analysis, data engineering, deep learning, and other skills I have related to communication skills like public speaking, networking, collaboration, and leadership.

TOOLS

Slide your mouse here

Git and Github
Google Colab
VS Code
Jetbrains Product
Tableau and Power BI
Microsoft Office
Jupyter Notebook
Google Office

TOOLS

Slide your mouse here

Some of the tools that I use are tools that are prioritized for data management.

OTHER SKILLS

Slide your mouse here

Mathematics
Google Cloud Platform
Microsoft Azure
Apache Hadoop
Apache Cassandra
Apache Spark
Apache HBase

OTHER SKILLS

Slide your mouse here

Some of the other skills I have are related to my experience in the field.