Skip to content
Sabid Mahmud, home

Open to software and applied ML roles

Sabid Mahmud · Software Engineer · Applied ML

I build web apps and deploy AI models.

I built and maintain Messio, a shared-kitchen app used by 10+ people, and PapLens, which runs our cervical-cytology deep learning models on self-hosted Kubernetes.

01

Full-stack development

TypeScript · Python · Next.js · SQL · MongoDB · REST APIs

02

Deep learning & computer vision

CNN · Transformers · PyTorch · FastAPI

03

Systems & deployment

Linux · Docker · Kubernetes · Gitops

01 / Selected work

From application to deployment.

More project details ↗

I build the interfaces, backend services, and deployment workflows behind a shared-kitchen product and a computer-vision research tool.

Messio member dashboard with daily meal choices, balance, and upcoming meal plans

Screenshot 1 of 7. Member dashboard

Full-stack product

Messio

A shared-kitchen application that turns meal records, deposits, and expenses into member balances and monthly statements.

My role · Design, development, deployment & maintenance (solo)

Started October 2025; in use since November 2025. Now used by 10+ people.

  • TypeScript
  • Next.js
  • MongoDB
Visit live app View project
PapLens evidence map comparing the original microscopy field with the model's patch-attention overlay

Screenshot 1 of 4. Attention map comparison

Applied ML · Research prototype

PapLens

PapLens runs cervical-cytology deep learning models from our research and lets people inspect the resulting attention maps in the original field.

Research prototype. Not for clinical or diagnostic use.

My role · Inference service, web application & deployment (solo)

About 4 s average inference for a standard input image. Self-hosted on K3s with GitOps deployment through Flux CD.

  • PyTorch
  • FastAPI
  • Next.js
  • K3s
View project

02 / Research

From microscopy
to inspectable models.

PapSmear Dataset · Co-curated dataset

Expert-annotated images
2,253
Source slides
300+

BRAC University · Undergraduate thesis

Explainable computer vision for cervical cytology.

With the research team, I co-curated an expert-annotated Pap-smear dataset and contributed to an attention-based deep learning classification pipeline using full-resolution image patches. I developed PapLens so people can run the deep learning models from our research and inspect the resulting attention maps in the original microscopy field.

Research prototype. Not for clinical or diagnostic use.

03 / Beyond the code

Building tools.
Sharing curiosity.

As Vice President of Brahmanbaria Science Club, I organize workshops and science fairs for high-school students, coordinating volunteers and local institutions to bring STEM activities to the community.

I graduated in Computer Science from BRAC University. I want to keep building software that people use in their daily work, from shared household tools to applications that make research easier to inspect.

More about me

Have something in mind?

Let’s build something useful.

Open to software engineering roles and projects involving web applications, research tools, and applied ML.

Dhaka, Bangladesh · UTC+06:00

Email me