Open to PhD positions & SWE roles

Sudarshan
Mondal.

Backend Engineer / Researcher

I build scalable backend and distributed systems for production. Curious about how things work and can be used for people.

View research View projects

01 About

An engineer who reads code closely | And a researcher drawn to machine learning and data.

I studied Computer Science & Engineering at NWU, where learned and practiced the fundamentals of Computer Science and Software Engineer. My research work was to read Bengali handwritten characters from image using CNN.

Can a system learn from messy, real-world data well enough to be trusted in production? That question is what pulls me from engineering toward research.

Since then I’ve worked across the stack in production teams — a distributed logging pipeline handling a million entries a day, a secure file-serving platform on AWS, high-throughput preprocessing services, and Android apps with real users. Most of that work is the slow, careful business of understanding code that already exists.

My research interest is machine learning and data science applied to software engineering: building and evaluating models that learn from real-world data and hold up in production. It sits exactly where my two threads meet - the rigour of academic evaluation on one side, the messy reality of production codebases on the other.

Distributed systemsSoftware engineeringApplied MLCloud & DevOps

02 Research & Publications

Work that holds up under review.

A peer-reviewed publication and the research direction I’m pursuing toward a PhD. Citation in standard format; abstract and links on expand.

Springer LNCS · ICICCT 2020 2020 Conference Paper

Convolutional Neural Networks Based Bengali Handwritten Character Recognition

Sudarshan Mondal, Nagib Mahfuz — International Conference on Innovative Computing and Communications (ICICCT 2020)

Mondal, S., & Mahfuz, N. (2020). “Convolutional Neural Networks Based Bengali Handwritten Character Recognition.” In: Innovative Computing and Communications, Lecture Notes in Computer Science, Springer (ICICCT 2020). https://doi.org/10.1007/978-3-030-52856-0_57

SpringerLink DOI

Handwritten character recognition for Bengali is challenging due to a large alphabet, compound graphemes, and high writer variability. This work presents a convolutional neural network trained on a dataset of isolated Bengali characters, with preprocessing and augmentation tailored to script-specific stroke structure, and reports classification performance competitive with prior approaches — offering a baseline for downstream document-analysis pipelines. (Summary — see SpringerLink for the published abstract.)

Looking for

PhD opportunity

The research direction I’m pursuing toward a PhD: machine learning and data science applied to real-world software systems — building models that learn from production-scale data and evaluating them honestly, building on my background in applied ML, software engineering, and cloud infrastructure.

03 Selected Projects

Things I’ve built end to end.

A filtered selection across backend, web, mobile, and ML. Filter by category below.

Backend · WebApp · 2026

Uni-Assis

A web app to find professors for PhD and MS research from top universities, around the globe.

NextJSPostgreSQLTailwind

Live

Backend · Distributed · 2024

Server Monitor

A distributed server-monitoring system built on microservices, with real-time alerting and fault isolation under high traffic.

PythonGolangRedisPostgreSQL

Web · Realtime · 2021

The FlashMob

A cross-platform short-video app with real-time chat, Celery-driven feed generation, and local video caching.

AndroidDjangoCelery

Demo

Web · Discovery · 2020

The Foodies

A personalised food & restaurant finder with smart search, photo-based reviews, and relevance-ranked suggestions. Plus tools for restaurants to manage menus and feature dishes.

AndroidDjangoPostgreSQL

Demo

Mobile · Education · 2021

Alphabet to Art

A kid-friendly Android drawing app with 120+ guided alphabet-art tutorials and an interactive canvas - custom pens, saved artwork, and learning through hands-on creativity.

AndroidJavaPaint

Demo

ML · Published · 2020

Bengali Handwritten Character Recognition

A convolutional neural network for recognising isolated Bengali handwritten characters. My undergraduate thesis, later published at ICICCT 2020 (Springer LNCS).

TensorFlowKerasPython

Paper Source

04 Experience

Where I’ve shipped.

Roles in production teams — framed by impact, not job descriptions.

Aug 2022 — Present

Jukto Digital

Khulna, BD

Software Engineer

Python · Django · AWS · Terraform · MySQL

Built a distributed logging system processing 1M+ log entries/day, improving data-integrity monitoring across teams.
Designed a secure, key-based file-serving system on AWS + Terraform with caching and audit logging.
Shipped a high-throughput preprocessing pipeline handling 5K+ requests/min, increasing API responsiveness by 65%.
Optimised 50+ complex MySQL queries over 10M+ records, and introduced service-repository patterns & code reviews that cut post-release bugs 15%.

Mar 2022 — Jul 2022

Angular Esports

Dhaka, BD

Software Engineer

Python · Django · Google Vision

Automated matchmaking and live scoring with a Django engine, cutting manual work by 35%.
Replaced manual data entry with a Django + Google Vision pipeline, improving accuracy and cutting input time by 30%.

Jul 2020 — Aug 2021

Appsinception

Khulna, BD

Software Engineer (Android)

Java · Firebase · Room · Retrofit

Built two production Android apps (News, Car Marketplace) with broad cross-device compatibility.
Refactored a legacy app to MVVM + repository architecture, improving performance 30% and retention 20%.
Drove crash rate from 45% to under 10% through systematic diagnostics and QA.

Feb 2020 — Jul 2020

Hellotask

Dhaka, BD

Data Scientist

Python · Django · Chart.js

Built an AI-powered recommendation engine, lifting sales 20% and retention 15%.
Automated analytical reporting with a Django + Chart.js dashboard with periodic auto-refresh.

05 Skills & Tools

The toolkit, grouped honestly.

No percentage bars — just what I reach for, organised by where it lives in the stack.

01

Languages & Frameworks

PythonJavaGolangDjangoFastAPIDRFCeleryPytestBash / Linux

02

Data & ML

TensorFlowKerasscikit-learnNumPyPandasMatplotlibFeature EngineeringModel Evaluation

03

Databases & Caching

PostgreSQLMySQLMongoDBRedis

04

DevOps, Cloud & Architecture

DockerAWSTerraformGitHub ActionsNGINXCI/CDMicroservicesSystem Design

06 Writing

Writing, notes & problem-solving.

Long-form essays on Medium, a growing archive of algorithm write-ups, and a garden of technical notes I keep as I learn.

May 2025 Networking

What is a Socket? A developer’s guide, explained like you’re 5 (with code)

A from-scratch walk through what sockets really are and how they work — with code — that grew out of trying to understand how DNS works under the hood.

Medium

May 2025 Distributed

Distributed scheduling gone wrong: the Celery Beat trap (and how we escaped)

A real failure mode every distributed system eventually hits — how duplicate scheduled tasks crept in, how we diagnosed it, and the fix that got us out.

Medium

May 2025 Architecture