Abrar Morshed AbrarSoul

About

I am a Project Researcher at GPT-Lab, Tampere University, working on AI agentic frameworks and generative AI for software engineering. My current work includes autonomous agents for code security, LLM-powered research tools, and teaching software process practices with GitHub and GitLab.

I hold an M.Sc. in Software, Web & Cloud from Tampere University, with four years of prior experience in MIS and backend development. I care about research that ships: systems that are useful, evaluable, and maintainable.

class AbrarMorshed:
    role = "Project Researcher @ GPT-Lab"
    affiliation = "Tampere University, Finland"
    focus = ["AI Agents", "Generative AI", "NLP", "Software Engineering"]
    email = "abrar.morshed@tuni.fi"

Now

Researching agentic systems for software engineering at GPT-Lab
Building tools around code security, retrieval, and LLM workflows
Open to research collaboration and conversation

Toolkit

_{Python · Java · JavaScript · TypeScript · C++ · R · PyTorch · React · Flask · Node.js · SQL · Docker}

Selected work

Agentic Code Security Auto-Fixer An autonomous agent that detects vulnerabilities and proposes repairs in source code — part of my GPT-Lab research on agentic software engineering. `Python` · `AI Agents` · `Security`	AI Supplier Contract Analyzer NLP pipeline for reading supplier contracts, extracting obligations, and surfacing risk language for faster review. `Python` · `NLP` · `LLMs`
Biological Aging Prediction Machine learning models for biological aging and frailty prediction, connecting health data to practical clinical signals. `Python` · `ML` · `Healthcare`	Agentic Weather Assistant A tool-using weather agent: natural-language queries, live API calls, and structured answers rather than a static dashboard. `Python` · `Agents` · `APIs`
Abstract Finder Research helper for locating and organizing paper abstracts, built to speed up literature review workflows. `Python` · `NLP` · `Research`	Content Generation with LLMs Experiments in controllable text generation with large language models — prompting, structure, and evaluation. `Python` · `LLMs` · `NLP`