Fahad (Md Golam Tawhid Fahad) is a Computer Science graduate from BRAC University and a software engineer based in Dhaka, Bangladesh. He works as Founding Engineer at ClassTablet and Junior Software Engineer at Nyntax, with experience in full-stack development, AI/ML, and computer vision research.

What does Fahad specialize in?

Fahad specializes in full-stack engineering (Next.js, NestJS, Django, Flask), backend architecture, TypeScript monorepos, computer vision, visual speech recognition, and biomedical ML. He builds multi-tenant SaaS platforms, secure web applications, and research-driven AI systems.

Where does Fahad work?

Fahad is Founding Engineer at ClassTablet (multi-tenant education platform, hybrid, Dhaka) and Junior Software Engineer at Nyntax (on-site, Dhaka). He previously conducted research at BIOSE, BRAC University (June 2024 – December 2025).

What is Fahad's education?

Fahad earned a B.Sc. in Computer Science from BRAC University (January 2022 – January 2026) with a CGPA of 3.70. He also completed HSC in Science from Amrita Lal Dey College, Barisal (GPA 5.00).

What research has Fahad done?

Fahad's research includes an undergraduate thesis on self-supervised multilingual visual speech recognition (Conformer-based VSR), a systematic review of vision-based fall detection, and a multimodal framework for stress detection using wearables and computer vision at BIOSE, BRAC University.

How can I contact Fahad?

You can contact Fahad via email at tawhidfahad199@gmail.com, LinkedIn at linkedin.com/in/g-t-fahad, or GitHub at github.com/Golam-Tawhid. His portfolio contact form is available at gtfahad.tech/#contact.

Is Fahad open to job opportunities?

Yes. Fahad is open to software engineering and machine learning roles. His focus areas include full-stack development, backend systems, computer vision, and applied AI for production products.

Shipping Payments and Messaging in a TypeScript Monorepo

Payments and notifications are where "it works on my machine" dies in production. Here's how we structured these subsystems at ClassTablet.

Payments & billing

Billing workflows are tenant-scoped: each organization has its own branding, pricing context, and parent payment views. Long-running billing tasks run through BullMQ so HTTP requests stay fast and failures can retry with backoff.

Messaging architecture

Web push: handled on the API with the web-push library
Email, WhatsApp, SMS: outbound adapters behind a unified notification service
Localization: multi-locale copy stored alongside templates

Observability

Sentry traces errors across the monorepo. Structured logs include tenant IDs (never raw PII) so we can diagnose failures per organization without exposing sensitive data in log aggregators.

Takeaway

Queue everything that talks to a third party. Your users' submit buttons will thank you.