Nazmin Hussain

backend & systems engineer — new delhi, india
nhussain.nnn@gmail.com
+91 7005823836
github.com/plastic07
01 — about

I am a software engineer

recent cs grad, currently working across backend infrastructure, distributed data pipelines, and applied machine learning.

languages
Python, C, C++, JavaScript
frontend & backend
React, Next.js, REST APIs, Django REST Framework, FastAPI, Flask, Supabase
databases
PostgreSQL, MySQL, SQLite, MongoDB, relational & NoSQL design
cloud & tools
GCP, AWS, Docker, Git, Linux, Bash, Pytest, Postman, Bruno
software engineering
OOP, SDLC, Agile, data structures & algorithms
AI / ML
NumPy, Pandas, PyTorch, LLM integration, RAG, embedding models, prompt engineering
02 — experience

Where I've worked

Software Developer Intern

Al Yusr International LLC · Muscat, Oman
  • Built an AI customer support platform for WhatsApp API, with RESTful analytics endpoints and an AI routing system featuring session management and loop prevention.
  • Built distributed webhook ingestion and normalization pipelines handling text, image, audio, video, and document events, with validation and error handling at scale.
  • Designed normalized PostgreSQL schemas and integrated LLM-based intent classification with structured JSON validation and rule-based fallbacks.

Data Analyst Intern

Firex Technologies and Services · New Delhi, India
  • Queried and analyzed large business datasets with SQL, building dashboards and automated reports on AWS to support decision-making.
  • Improved data pipeline reliability through preprocessing and quality-validation checks.

Software Developer Intern

Find Your N · New Delhi, India
  • Developed scalable full-stack services with Python, Supabase, and React, following modular, object-oriented backend architecture.
  • Architected authentication, storage, and frontend workflows using REST API design principles for client-server communication.
  • Integrated third-party APIs and enforced Git-based workflows: branching, pull requests, code review.
03 — projects

Things I've built on my own

Minimal Secure VPN System Using TUN Device

A lightweight VPN built from scratch using Linux TUN interfaces and UDP sockets, with AES-256-GCM authenticated encryption via OpenSSL for secure point-to-point tunneling.

C · Python · OpenSSL · UDP Sockets · TUN Interface · Linux · Tailscale

Order Management System — Django REST API with Google OAuth 2.0

A secure Django REST backend with Google OAuth 2.0 authentication, access/refresh token handling, and filtered, role-based query access across RESTful endpoints.

Python · Django · Django REST Framework · Google OAuth 2.0 · SQLite

WhatsApp Automation Platform

An AI-powered WhatsApp Business automation backend that classifies incoming messages with an LLM, sends smart auto-replies, and routes technical issues to a human support team, with an operations dashboard for real-time analytics.

Node.js · Express.js · PostgreSQL · LLM API · Meta WhatsApp Cloud API

Geospatial Sensor Data Acquisition Platform

A cross-platform desktop application for real-time camera, GPS, and IMU data acquisition with synchronized timestamping and interactive sensor visualization, built on a multithreaded, event-driven architecture integrating device SDKs, serial communication, and OpenCV pipelines for responsive, low-latency processing. Supports data logging, playback, and export of synchronized sensor streams to CSV/JSON.

C++ · Qt6 · OpenCV · QtSerialPort · CMake · OpenGL · Git · Linux

Mini ETL Pipeline for Data Warehouse

An end-to-end ETL pipeline extracting JSON data from the AniList REST API and loading it into a PostgreSQL data warehouse, simulating production data workflows. Handles ingestion, transformation, deduplication, normalization, and schema enforcement, with relational schemas and SQL queries for filtering, joins, and aggregations to produce analytics-ready datasets for KPI reporting.

Python · SQL · PySpark · Pandas · PostgreSQL · JSON · REST APIs

Autonomous Multi-Agent Outreach Platform

A multi-agent system built with CrewAI that automates company research, resume analysis, candidate-job matching, contact discovery, and personalized outreach. Specialized AI agents are orchestrated for web scraping, semantic search, task automation, and sequential decision-making, with LLM-powered pipelines handling prompt engineering, structured output parsing, and automated email delivery.

Python · CrewAI · Semantic Search · REST APIs · Git

Rubik's Cube Solver

A work-in-progress Rubik's cube solver written in Python, exploring cube-state representation and solving algorithms from scratch.

Python
04 — research & publications

Published work

best paper

Compiler Optimization Using Machine Learning

International Conference on ICT for Digital, Smart and Sustainable Development (ICIDSSD '26), Scopus-indexed, New Delhi. April 2026.

Surveyed the MLGO, CHAMELEON, and DeepTune frameworks, analyzing supervised, reinforcement, and deep learning methods for compiler phase ordering and heuristic tuning.

accepted

Diagnosis from Tissue: Deep Learning for Breast Cancer Classification

International Conference on Artificial Intelligence and Simulation (CAIS 2026), Kuala Lumpur, Malaysia. September 2026.

Patient-wise comparative evaluation of ResNet50, MobileNet, and EfficientNet for breast cancer histopathology classification on the BreaKHis dataset, identifying EfficientNet as offering the strongest accuracy-efficiency trade-off. Co-authored with Monazzah Shaheen, Sania Bushra, and Seema Insha Parveen. To be published via conference proceedings and indexed in IEEE Xplore, EI Compendex, and Scopus.

05 — education

Jamia Hamdard

B.Tech, Computer Science Engineering

New Delhi, India
  • CGPA: 9.04 / 10.0