About Us · The Lab

Who we
are.

A student-led AI research group focused on medical imaging, computer vision, and responsible machine learning.

Affiliation
AIUB · Dhaka
Discipline
AI Research
Focus
Medical Imaging · NLP
Approach
Open · Rigorous

Founded in 2024, Arché Intelligence Lab is a research-focused academic initiative dedicated to advancing artificial intelligence through innovation, scientific exploration, and real-world impact.

The lab works across areas such as machine learning, medical imaging, computer vision, and explainable AI — bringing together emerging technologies and collaborative research to build responsible, intelligent, and future-ready solutions.

The lab promotes rigorous research, open science, and impactful publications. Our goal is to contribute credible knowledge to the global academic community — work that holds up after the dataset changes and the spotlight moves on.

/ Mission

To advance AI research that is transparent, ethical, and impactful — contributing credible knowledge to the global academic community.

— Arché Intelligence Lab · 2024
Principles

Core values.

The beliefs that guide every project and publication — a working method we hold each other to, not a poster on a wall.

01

Transparency

/ open science

Open science principles guide every project. Our methods, data, and results are shared with the community — not gated, not paraphrased, not after the embargo lifts.

02

Data-Driven Methodology

/ evidence first

Every claim is backed by rigorous experimentation and statistical validation. Effect sizes, confidence intervals, and ablations — not just the headline number on the median split.

03

Ethical Research

/ responsibility

We prioritize responsible AI development with careful consideration of societal impact — who the system helps, who it could harm, and how that balance shifts when it leaves the lab.

04

Explainability

/ interpretability

AI systems should be interpretable. We build models whose decisions can be understood and trusted — by clinicians, by reviewers, and by the people the predictions are made about.

05

Reproducibility

/ scientific integrity

Our experiments are designed to be fully reproducible — code, configuration, and seeds. If a result can't be re-run by someone else, it isn't a result yet.

Collaborate

Curious about a project, or want to work with us?

We welcome motivated students and researchers who share our commitment to rigorous, impactful AI work.