R
Portrait of Rezab Ud Dawla

Rezab Ud Dawla

I am a software engineer and researcher interested in reliable and maintainable LLM applications, software engineering for AI-enabled systems, data management, data quality, privacy, and text mining. Currently, I work as a Software Engineer at Linno, developing secure full-stack features and REST APIs and building a self-hosted RAG support assistant with FastAPI, llama.cpp, and hybrid retrieval.

My academic research spans attention-based scene graph generation and imbalanced-data classification. My undergraduate education at United International University focused on Computer Science and Engineering, with a Data Science major; I graduated with a CGPA of 3.72/4.00. I have co-authored three peer-reviewed conference publications at SKIMA, IEEE TENCON, and MIET. Previously, I worked at ShareTrip and RedDot Digital Limited and served as an Undergraduate Teaching Assistant.

Publications

  1. Methodology figure for Attention-Based Scene Graph Generation: A Review

    Afsana Airin, Rezab Ud Dawla, Ahmed Shabab Noor, Muhib Al Hasan, Ahmed Rafi Hasan, Akib Zaman, and Dewan Md. Farid.

    Attention-Based Scene Graph Generation: A Review

    14th International Conference on Software, Knowledge, Information Management and Applications, 2022.

    DOI · PDF
  2. Methodology figure for A Novel Method for Imbalanced Data Classification Based on Label Reassignment

    Ahmed Shabab Noor, Afsana Airin, Rezab Ud Dawla, Ahmed Rafi Hasan, Muhib Al Hasan, Akib Zaman, and Dewan Md. Farid.

    A Novel Method for Imbalanced Data Classification Based on Label Reassignment

    TENCON 2022 — IEEE Region 10 Conference, Hong Kong, 2022.

    IEEE Xplore · PDF
  3. Methodology figure for The Impact of Data Locality on the Performance of Cluster-Based Under-Sampling

    Ahmed Shabab Noor, Muhib Al Hasan, Ahmed Rafi Hasan, Rezab Ud Dawla, Afsana Airin, Akib Zaman, and Dewan Md. Farid.

    The Impact of Data Locality on the Performance of Cluster-Based Under-Sampling

    Machine Intelligence and Emerging Technologies, LNICST 491, Springer, 2023.

    DOI · PDF

Undergraduate Research Experience

Final Year Design Project — Improvement of Scene Understanding Using Scene Graphs

February 2022 – October 2022 · United International University

Advised by Professor Dr. Dewan Md. Farid, with Akib Zaman as co-advisor.

  • Conducted the gap analysis and contributed to the literature review of attention-based scene graph generation methods.
  • Contributed to the project design and implementation of an attention-based scene graph generation system for object-relationship prediction.
  • The project used Python, PyTorch, NumPy, and OpenCV. In a preliminary test, the model correctly classified predicates in 13 of 15 images (86.67%). Recall@K was specified for future evaluation.

Experience

  1. Software Engineer — Linno

    January 2026 – Present (Dhaka, Bangladesh)

    I develop secure full-stack features and REST APIs with PHP and React, using role-based access control, optimized SQL queries, and Redis caching. I also built a self-hosted RAG support assistant with FastAPI, llama.cpp, hybrid retrieval, and confidence-based human escalation.

    Notable project:AI Troubleshooting Assistant

  2. Software Engineer — ShareTrip

    July 2024 – January 2026 (Dhaka, Bangladesh)

    I built Python and FastAPI services and ETL workflows for an internal analytics platform. I improved high-volume booking systems through Redis caching, background job queues, and containerized deployments.

    Notable projects:Single Sign-On PlatformReporting and Analytics Management System

  3. Software Engineer Intern — RedDot Digital Limited

    October 2023 – February 2024 (Dhaka, Bangladesh)

    I developed and tested Laravel features, including PHPUnit coverage, across the full software-development lifecycle within an Agile team.

  4. Undergraduate Teaching Assistant — United International University

    February 2023 – May 2023 (Dhaka, Bangladesh)

    Taught CSE 4165: Web Programming to approximately 45 students, assessed course learning outcomes, and supervised student software projects.

Selected Projects

Architecture diagram for the AI Troubleshooting Assistant

AI Troubleshooting Assistant

Industry work · March 2026 – June 2026

Built and deployed a RAG pipeline with a quantized, self-hosted small language model for first-line technical support. Combined semantic and keyword retrieval, confidence-based human escalation, and evaluation on labeled historical conversations.

Python · llama.cpp · FastAPI · Vector DB · HelpScout API

Architecture diagram for the Single Sign-On Platform

Single Sign-On Platform

Industry work · November 2024 – May 2025

Built a centralized SSO platform with Redis-backed session sharing, role- and service-based access control, event tracking, and secure management APIs.

Yii · MySQL · Redis · RabbitMQ · Swagger UI

Architecture diagram for the Reporting and Analytics Management System

Reporting and Analytics Management System

Industry work · September 2025 – December 2025

Developed authenticated FastAPI endpoints, ETL pipelines, and Streamlit dashboards, with Redis caching and Docker-based deployments.

Python · FastAPI · Streamlit · MySQL · Redis · Docker

Rock detection model and sample classification results

Rock Detection Model

UIU Mars Rover Team · 2022

Built a binary vision model on a custom dataset for classifying fossil and metamorphic rocks. Applied VGG16 feature extraction and increased reported accuracy from 82% to 97%.

Python · Pandas · OpenCV · TensorFlow · VGG16

Honors & Rewards

University Scholarship Recipient

Received nine merit-based scholarships in recognition of consistent academic excellence across individual semesters and overall academic performance, with each award covering either 25% or 50% of tuition.

1st Runner-Up - CSE Project Showcase

Team Lead · Intra-university project show competition · Fall 2022

Led Team Firefox in developing and presenting the UIU Club Management System. The platform centralized club membership records, event scheduling, and activity tracking to simplify administrative work and encourage stronger student participation.

Certifications

Others

Currently Reading

Cover of Illustrated Guide to AI by The Welch Labs
Illustrated Guide to AIThe Welch Labs
Cover of Wool by Hugh Howey
WoolHugh Howey