Hi! I am Abu Taher, a doctoral researcher at the University of Oulu, Finland. I work at the intersection of distributed systems, edge AI and healthcare. My research focuses on envisioning and implementing architecture that will power hospitals of the future.
If you want to talk with me, feel free to reach out either by email or through LinkedIn. Read More
Working on the edge-cloud continuum, studying geo-distributed multi-tier, multi-cluster critical medical application orchestration.
Skills: Docker · Docker Swarm · Kubernetes · Linux OS · Cloud Computing (AWS and CSC)
I got the interview opportunity at Huawei through the campus recruitment exam, where more than 200 students participated and only 5 were finally selected. In the first week of my job, I went through an intensive workshop, learning about the company’s culture, values, business practices, and the sheer scale of its operations in Bangladesh. During the following three weeks of my one-month job, I studied the network topology of Robi, one of the biggest telecom operators in Bangladesh, along with Huawei’s CloudEngine equipment that powers Robi’s daily operations at a massive scale. During this time, I gained knowledge about 4G IMS, network virtualization, company values and culture, and also had the opportunity to network with colleagues.
Skills: Voice Core · 4G IMS · Software Defined Networking · Network cloud virtualization · Network Engineering · Local Area Network (LAN) · Wide Area Network (WAN)
Docker | Kubernetes | Linux OS (Ubuntu) | GitLab | Redis | Git | Prometheus | Grafana
C, C++ | Basic Python | JS/HTML/CSS basics | Keras Library | Microsoft SQL Server, Data structure and Algorithms, | Programming with ASP.NET
TensorFlow | Keras | Scikit-learn | Dlib | OpenCV | face_recognition | NLTK | SciPy | NumPy | Pandas
Wireshark | Cisco packet tracer
R | MatLab | CST studio suite(antenna design software) | Microwind DSCH2
Microsft SQL server | MySQL
As part of my Distributed Systems course (MSc), I built this full-stack chatbot application by using GPT-2 (124M parameters) as a base model. The GPT-2 model was split layer-wise into two halves and deployed on two local PCs during development. These two parts communicate over the gRPC protocol. To make the application fully functional and user-friendly, I developed 7 containerized microservices: UI service, API gateway, authentication, database, inference, monitoring, prompt engineering, and load balancing. These services interact through RESTful APIs to provide a seamless user experience. For the backend, I used MySQL for storing user and chat history data. Part of this project was containerized using Docker and successfully deployed on the AWS (t2.medium EC2) cloud. I performed unit tests, load tests, and chaos engineering (98.5% uptime) to evaluate robustness and performance. From this project, I learned to design, develop, test, and deploy a cloud-native application.
Developed an intelligent anomaly detection device using Raspberry pi and Edge AI. Implemented IoT communication pipeline using MQTT and Node-RED, which demonstrates an understanding of network communication protocols and data transmission.
Technologies: Raspberry Pi Pico W, MQTT, InfluxDB, Grafana, Random Forest
Cotton Leaves Disease Classification Using Custom Convolutional Neural Network (CNN) Model
[ 2023 ]
A. Taher and F. Hossain, ‘Cotton Leaves Disease Classification Using Custom Convolutional Neural Network (CNN) Model’, presented at the INTERNATIONAL CONFERENCE ON NEXT-GENERATION COMPUTING, IoT AND MACHINE LEARNING NCIM 2023, Department of Computer Science and Engineering of Dhaka University of Engineering & Technology, Gazipur, Bangladesh: IEEE.
[ 2023 ]
Hasan, R., Ullah, S. M. A., Nandi, A., & Taher, A. (2023, September). Improving Pneumonia Diagnosis: A Deep Transfer Learning CNN Ensemble Approach for Accurate Chest X-Ray Image Analysis. In 202 3 International Conference on Information and Communication Technology for Sustainable Development (ICICT4SD) (pp. 109-113). IEEE.
[ 01 Sept 2024 – 31 December 2025 ]
University of Oulu, Finland
Specialization: Artificial Intelligence
Courses: Natural Language Processing and Text Mining, Towards Data Mining, Introduction to Optimization, Biosignal Processing 1, Machine Vision, Machine Learning, Big Data Processing and Applications, Computer Graphics, Affective Computing, Multi-Modal Data Fusion, Deep Learning, Distributed System, Multiprocessor Programming, Internet of Things, Human AI Creativity and C++ Programming
[ 27 Nov 2017 – 20 Mar 2023 ]
Khulna University of Engineering & Technology
Thesis: Cotton Leaves Disease Classification Using Custom Convolutional Neural Network (CNN) Model
[ 30 Apr 2015 – 23 Jul 2017 ]
Government Victoria College Narail
Final grade: 5.00 out of 5.00
[ 1 Jan 2013 – 30 May 2015 ]
Narail Government High School
Final grade: 5.00 out of 5.00
Foundations: Data, Data Everywhere
Authorized by Google and offered through Coursera
SQL Server Developer Career Track
Authorized by DataCamp
General Scholarship (SSC)
Awarding institution: Jessore Board
Technical Scholarship
Awarding institution: Khulna University of Engineering and Technology
Presentation Workshop
KUET Career Club
Line-follower Robot
Manipulation of Electrons (MEC), KUET
Printed Circuit Board Design
Dept. of ECE, KUET
Advisor [2023-Current]
Narail Student Association of KUET
General Member [2018–2019]
KUET Career Club
“I was Abu Taher’s supervisor and course teacher. He was one of the most active and passionate students in his class. Being his thesis supervisor in his senior year, I mentored him closely and came to know about his diligent nature in his research work. Abu Taher developed a robust CNN model capable of accurately classifying various cotton leaf diseases during his bachelor’s thesis, demonstrating superior performance compared to other CNN models. He was very rigorous in his research. I believe that he will continue to excel in his future studies and contribute significantly to any research endeavors he undertakes.”
“Abu Taher is a dedicated, highly energetic student. As a project supervisor and course teacher, I have observed Taher’s academic integrity and intellectual ability very closely. During the electronics project, he developed a patient monitoring system using a pulse oximeter, ESP Arduino, and thermistor, which could send the data to an Android app. In my class, he was always disciplined, responsive, and obedient throughout the course. He has always tried to achieve a conceptual understanding of every course by analyzing the theory and applying the theoretical knowledge in practical work, which helped him to be awarded an above-average GPA in all of my courses.”