All Skills

My Projects and the Tech Stack

These are the tools and technologies that I use daily to develop web applications. With a strong foundation in frontend, I focus on delivering clean, maintainable code and intuitive user experiences.

Agile Methodology

Agile Methodology

FastAPI

FastAPI

PostgreSQL

PostgreSQL

Apache Kafka

Apache Kafka

MQTT (HiveMQ)

MQTT (HiveMQ)

Redis

Redis

Docker

Docker

Kubernetes

Kubernetes

React

React

TypeScript

TypeScript

VHDL

VHDL

FPGA

FPGA

Xilinx Vivado

Xilinx Vivado

C

C

Raspberry Pi Pico

Raspberry Pi Pico

TinyUSB

TinyUSB

Django

Django

Python

Python

SQLite

SQLite

SQLAlcehmy

SQLAlcehmy

Tailwind CSS

Tailwind CSS

Flask

Flask

Git

Git

GitHub

GitHub

Prometheus

Prometheus

Grafana

Grafana

Kubernetes

Kubernetes

Istio

Istio

PROJECTS

OnTime – Real-Time Public Transport Tracking Platform

Completed on May 19, 2026

Category: Academic
FastAPI
PostgreSQL
Apache Kafka
MQTT (HiveMQ)
Redis
Docker
Kubernetes

Contributed as a Backend and DevOps developer to a team-built, event-driven public transport tracking platform, working across route and fleet-management microservices, PostGIS-backed geospatial data ingestion, GPS telemetry simulation using MQTT and Kafka, Redis-backed WebSocket updates, authentication, API gateway integration, automated testing, CI/CD, container publishing, security, and Kubernetes deployment.

Kandypack – Logistics Management System

Completed on Nov. 1, 2025

Category: Academic
FastAPI
PostgreSQL
React
TypeScript

Contributed as a full-stack developer to a rail-and-road logistics management system, implementing order allocation across trains and destination stores with quantity and capacity validation, train-arrival and store-inventory workflows, and delivery rostering based on routes, available trucks, and eligible staff. Integrated FastAPI services with MySQL stored procedures for order processing, employee selection, train arrivals, and delivery assignments, while contributing authenticated store-specific order management, automated status updates, business reporting, and interactive dashboards.

Nano Processor

Completed on Aug. 11, 2025

Category: Academic
VHDL
FPGA
Xilinx Vivado

Designed and implemented a four-bit nano processor using VHDL, integrating components such as registers, an arithmetic and logic unit, a program counter, instruction decoding, and program memory.

USB MIDI Keyboard

Completed on Feb. 1, 2026

Category: Personal
C
Raspberry Pi Pico
TinyUSB

Converted a 61-key digital keyboard into a USB MIDI controller using a Raspberry Pi Pico. Reverse-engineered and scanned its 11×6 key matrix, implemented key-state detection and MIDI Note On/Off messaging in C, and configured TinyUSB USB descriptors for plug-and-play MIDI communication.

Django Portfolio CMS

Completed on Jan. 31, 2026

Category: Personal
Docker
Django
Python
SQLite
SQLAlcehmy
Tailwind CSS

Built a Django-based portfolio and blogging CMS with Django Admin management for articles, projects, skills, certificates, profile information, media, and resume content. Implemented rich-text publishing, search, tag filtering, pagination, automatic slugs, view tracking, and configurable SEO metadata, and containerized the application for deployment using Docker Compose, Gunicorn, and Caddy with automated TLS, persistent storage, and scheduled backups.

MovieNetR

Completed on Feb. 2, 2026

Category: Experimental
Python
SQLite
SQLAlcehmy

Built a movie recommendation network from Reddit discussions and IMDb data, using GLiNER-based zero-shot named entity recognition to extract movie mentions and RapidFuzz to match them against canonical IMDb titles. Constructed a weighted co-recommendation graph with NetworkX and persisted processed data using SQLAlchemy and SQLite, then developed an interactive Sigma.js explorer for searching films, navigating recommendation neighborhoods, and revealing clusters of related movies and bridge titles connecting different genres.

MovieNetR – Movie Recommendation Network

Completed on Feb. 17, 2026

Category: Experimental
Python
SQLite
SQLAlcehmy

Built a movie recommendation network from Reddit discussions and IMDb data, using GLiNER-based zero-shot named entity recognition to extract movie mentions and RapidFuzz to match them against canonical IMDb titles. Constructed a weighted co-recommendation graph with NetworkX and persisted processed data using SQLAlchemy and SQLite, then developed an interactive Sigma.js explorer for searching films, navigating recommendation neighborhoods, and revealing clusters of related movies and bridge titles connecting different genres