Kawsher HRidoy

ML & Systems developer Team Lead, Cortex Crew

AI systems.Built for the real world.

I lead Cortex Crew at Daffodil International University in Dhaka. My work pairs applied machine learning with careful engineering—from cybersecurity awareness to decision-support and infrastructure recovery.

Explore selected work Get in touch

CSAD 2026 Champion.
IEEE ICADHI 2026 1st Runners-up.
Four showcased systems.

Selected work

From evidence to action.

Four team-built systems. Different problems, the same care for what happens next.

Cybersecurity Victim awareness

Darktrace3

Makes an infostealer incident understandable—showing exposed accounts and sites without revealing passwords.

CSAD 2026 Champion

Python

Source private

Exposure, made understandable

Evidence flow

Exposure evidenceAccount & site evidence
Masked account/site viewCredentials remain masked
ReviewUnderstand the exposure

Victim awareness. Not credential disclosure.

Illustrative system diagram.

Read Darktrace3 project story

The problem

Malware and infostealer logs can leave people unsure which accounts were put at risk and what to do next.

What we built

Cortex Crew turned exposed-site and account evidence into a clear awareness view, with passwords kept masked throughout.

The ethical boundary

Victim awareness needs useful evidence, not unnecessary disclosure of sensitive credentials. The view communicates exposure while preserving credential privacy.

Health systems Patient ownership

Niro

Brings Bangla medical documents and a connected health history back to the patient—and useful context to their doctor.

IEEE ICADHI 2026 1st Runners-up

FastAPI
Bangla OCR

Source private

A connected history. A clearer visit.

Care workflow

Bangla documentsLab reports & prescriptions
Patient-owned timelineHistory across visits
Doctor-ready summaryContext for the consultation

Clinician in the loop. The AI does not give final medical advice.

Illustrative system diagram.

Read Niro project story

The problem

Lab reports, prescriptions, and medical history are difficult to keep together and interpret across visits.

What we built

Cortex Crew built a patient-owned record with Bangla document reading, time-series tracking, and an organised case summary for the doctor.

The clinical boundary

The AI helps organise context; it does not give final medical advice. The clinician makes the medical decision, with the patient’s history in view.

ML infrastructure Predictive recovery

Autopilot

Connects GPU failure prediction to a recovery action—giving running work a chance to move before a node fails.

AI Innovation Hackathon 2026 Finalist

Python
LightGBM
Kubernetes

View Autopilot source

Prediction with an operational purpose

GPU telemetryObserve the hardware
PredictionIdentify potential failure
Cordon nodePrevent new scheduling
Migrate jobMove running work

Approximately 100 seconds ahead: stated project behavior, not a benchmark verified here.

Illustrative system diagram.

Read Autopilot project story

The problem

GPU node failures interrupt valuable running work when the infrastructure has no time to react.

What we built

Cortex Crew built a telemetry-driven prediction loop that identifies potential failures about 100 seconds ahead, cordons the node, and migrates the running job. This describes the project’s stated behavior, not a guarantee of uptime.

The engineering decision

The prediction is tied to a concrete Kubernetes recovery action, so the signal has an operational use rather than stopping at an alert.

My contribution

Team lead—ML model, controller, and system integration, as documented in the public project repository.

Education systems Decision support

AI Mentor

Makes academic and attendance evidence usable for mentors, with transparent rules and an AI brief—not a black-box decision.

AI Project Competition 2026 Finalist

Next.js
TypeScript
Azure OpenAI

View AI Mentor source

Rules first. A person decides.

Eight inputsAcademic & attendance
Deterministic rulesAuditable ranking
Mentor briefAI explains the context
Human decisionMentor stays accountable

AI explains the evidence. It does not determine the ranking.

Illustrative system diagram.

Read AI Mentor project story

The problem

Mentors need to see who needs attention and why, without a black-box score becoming the decision.

What we built

Cortex Crew built a system that evaluates eight academic and attendance inputs using deterministic rules, then adds an AI-generated mentor brief.

The engineering decision

Ranking stays deterministic and auditable. AI explains context for the person who makes the decision; it does not invent or influence the score.

Competition record · 2026

Work put to the test.

A selected record of team results: one championship, one 1st Runners-up finish, and three finalist placements.

Jan 2026
Champion
CSAD 2026 Project Showcasing Competition
Cyber Security Centre, DIU
Darktrace3
Jun 2026
1st Runners-up
IEEE ICADHI 2026 Project Showcase
IEEE DIU Student Branch
Niro
Jul 2026
Finalist
AI Innovation Hackathon 2026
Department of CSE, DIU
Autopilot
Jul 2026
Finalist
AI Project Competition 2026
Department of CSE, DIU
AI Mentor
Jul 2026
Finalist
5th Data Science Summit
DIU
Niro

Engineering approach

The decisions behind the systems.

Deterministic where it counts
Use clear, inspectable logic for scoring, ranking, and actions that need to be defended.
AI explains; people decide
Put AI where it clarifies evidence and makes context usable, while keeping accountable humans in control.
Measure against the real baseline
Start with the failure, workflow, or user burden already present—and make the technical decision answer to it.

About

The model is only part of the work.

I’m Kawsher, an ML & Systems developer and Team Lead of Cortex Crew at Daffodil International University. I lead the team from early planning to a live, defensible competition demo.

I care about the system around the model: the evidence it uses, the interface people see, and the action it enables. My current direction includes competitive cybersecurity.

Dhaka, Bangladesh Cortex Crew, DIU

Tools I work with

Modelling

Python
LightGBM

Backend & data

FastAPI
PostgreSQL

Frontend

Next.js
TypeScript

Infrastructure

Docker
Kubernetes
Linux

AI services

Azure OpenAI

Contact

Let’s build
something useful.

Have a problem worth thinking through?

kawsher@hridoy.xyz

AI systems.Built for the real world.

From evidence to action.

Darktrace3

The problem

What we built

The ethical boundary

Niro

The problem

What we built

The clinical boundary

Autopilot

The problem

What we built

The engineering decision

My contribution

AI Mentor

The problem

What we built

The engineering decision

Work put to the test.

CSAD 2026 Project Showcasing Competition

IEEE ICADHI 2026 Project Showcase

AI Innovation Hackathon 2026

AI Project Competition 2026

5th Data Science Summit

The decisions behind the systems.

Deterministic where it counts

AI explains; people decide

Measure against the real baseline

The model is only part of the work.

Tools I work with

Modelling

Backend & data

Frontend

Infrastructure

AI services

Let’s buildsomething useful.

Let’s build
something useful.