Shadman Sakib

Dhaka, Bangladesh

I am a Software Engineer at Synesis IT PLC in Dhaka, Bangladesh, where I build backend systems in Java and Spring Boot for enterprise-scale government information platforms. I hold a B.Sc. in Software Engineering from the Islamic University of Technology (IUT), and my day-to-day work — maintaining legacy codebases, refactoring for structural consistency, and securing distributed microservices — sits close to the software evolution problems I care about as a researcher.

My research interests center on automated software verification and model conformance, large language models for requirements engineering, and trustworthy, reliable AI. My undergraduate thesis built an agentic, prompt-driven pipeline that generates structured functional requirements from unstructured app-store reviews, evaluating machine-generated artefacts against human baselines to surface systematic conformance gaps. This work was accepted at NLP4RE’26 @ REFSQ, and a related study on LLM reliability in clinical prediction appears in IEEE Xplore.

Feel free to reach out by email, or browse my CV, publications, and projects.

news

Feb 01, 2026	Our paper From Reviews to Requirements: Can LLMs Generate Human-Like User Stories? was accepted at NLP4RE’26 @ REFSQ!
Oct 01, 2025	Joined Synesis IT PLC as a Software Engineer I, working on backend systems for enterprise-scale government platforms.

selected publications

REFSQ
From Reviews to Requirements: Can LLMs Generate Human-Like User Stories?

Shadman Sakib and others

In Proceedings of the 9th Workshop on Natural Language Processing for Requirements Engineering (NLP4RE), co-located with REFSQ, 2026

Abs arXiv Bib HTML

An automated pipeline that generates structured software requirements from app-store user feedback. We evaluate zero-shot, few-shot, and chain-of-thought prompting for their effect on requirement completeness, consistency, and conformance to human-written baselines, and categorise the failure modes that arise in LLM-generated requirements.
@inproceedings{sakib2026reviews, title = {From Reviews to Requirements: Can LLMs Generate Human-Like User Stories?}, author = {Sakib, Shadman and others}, booktitle = {Proceedings of the 9th Workshop on Natural Language Processing for Requirements Engineering (NLP4RE), co-located with REFSQ}, year = {2026}, url = {https://arxiv.org/abs/2603.28163}, }
IEEE
From Chat to Checkup: Can Large Language Models Assist in Diabetes Prediction?

Shadman Sakib and others

In IEEE Xplore, 2025

Abs Bib HTML

We benchmark LLM-based inference against traditional machine-learning classifiers on the Pima Indians Diabetes Dataset, analysing the reliability, interpretability, and failure conditions of generative models in a safety-critical decision-support context.
@inproceedings{sakib2025diabetes, title = {From Chat to Checkup: Can Large Language Models Assist in Diabetes Prediction?}, author = {Sakib, Shadman and others}, booktitle = {IEEE Xplore}, year = {2025}, publisher = {IEEE}, url = {https://ieeexplore.ieee.org/document/11171691}, }