Ananta Akash Podder

anantaakash.podder@gmail.com

I am an experienced Software Engineer specializing in application development, with a strong background in Agile methodologies and Linux. Passionate about advancing industry practices through research, my interests include web development, machine learning, natural language processing and software security. As a dedicated team player, I am committed to continuously improving through collaboration and learning.

Research Experience

Bangla Topic Modeling Techniques and Comparison

Undergraduate Thesis

Supervised By: Mohammad Abdullah Al Mumin

Thesis Title: Bangla Topic Modeling Techniques & Comparison using News Trend Analysis

  • The research focused on developing a topic modeling algorithm for Bangla and using it to extract topics from online news documents (around 288k).
  • The findings of the research are a modified-LDA that works well on Bangla data and a method to find news trends.
  • Fed the Bangla dataset and tuned the hyper-parameters to fit the data and used it to analyze news trends in 9 different categories over the past 5 years.
June 2021

Publication

  • A Systematic Literature Review on English and Bangla Topic Modeling, Journal of Computer Science, 17(1), 1-18. 🔗

    • Education

    Bachelor of Science in Computer Science and Engineering

    Shahjalal University of Science and Technology, Sylhet, Bangladesh

    CGPA: 3.72 [out of 4.00]

    Jan 2017 - July 2021

    Higher Secondary Certificate

    Dhaka Residential Model College, Dhaka, Bangladesh

    GPA: 5.00 [out of 5.00]

    2014 - 2016

    Work Experience

    Tekarsh | MarginEdge🔗

    Software Engineer
    • Contributed to MarginEdge’s subscription billing system, supporting 10k+ USA restaurants, by provisioning cloud infrastructure via AWS CDK.
    • Integrated LaunchDarkly🔗 into the CDK project to introduce feature flag mechanism that improved release flexibility and ensured zero-downtime rollbacks during production.
    • Mentored junior developers, fostering knowledge-sharing and acquired proficiency in Agile work culture.
    Frameworks, Tools & OS:
    JavaSpring Boot, ReactJS, AWS CDK (EC2, Lambda, DynamoDB, SQS, SNS, S3, CloudFormation)
    2025 - Present

    Progoti Systems Limited | Tallykhata🔗

    Software Engineer
    Tallykhata Backend
    • Enhanced the external notification system by sending notifications in chunks using Celery and Celery Beat, resulting in a 70% performance improvement for a user base of over 5 million.
    • Helped to establish Robust CI/CD Pipelines in GitLab with Docker & Docker-compose, reducing deployment times, minimizing errors, and enhancing overall development efficiency.
    • Worked on syncing user data from Android Application to Django Backend Server.
    • Integrating third-party APIs and Deploying Django applications using Nginx & Gunicorn.
    • Developed PWA (Progressive Web App) for multiple dynamic URLs for feature called 'Credit Collection Link'.
    Frameworks, Tools & OS:
    Django, Docker, Linux, Celery, Nginx, Gunicorn
    Tallypay Backend
    • Horizontally scaled & containerized Redis Server using Redis Sentinel and HAProxy, deploying a cluster of master nodes and replicas, ensuring high availability & addressing single point of failure.
    • Implemented a publisher-subscriber architecture using Java Spring and RabbitMQ to facilitate eKYC information verification for Tallypay users by the TechOps team.
    • Developed API mechanisms to block/unblock any specific user debit or credit transaction to prevent fraud.
    Frameworks & Tools:
    Spring-Boot, HAProxy, Redis, Rabbit-MQ
    Tallykhata Mobile App
    • Worked on MVVM(Model, View, ViewModel) pattern in Android Application.
    • Worked with Android's Webview & JavaScript Interface to integrate various 3rd party payment system APIs.
    Frameworks & Tools:
    Android SDK, Retrofit
    2021 - 2025

    Skill

    Programming Languages & Essentials

    Java, Python, C/C++, JavaScript, TypeScript, SQL, HTML

    Application Frameworks

    Django, Spring-Boot, ReactJS, FastAPI, Android SDK, Laravel

    DevOps, Cloud and OS

    Linux OS, Docker, Kubernetes, Helm, AWS (EC2, Lambda, DynamoDB, SQS, SNS, S3, CloudFormation), Jenkins, Ansible

    Tools

    Rabbit-MQ, HAProxy, Nginx, Gunicorn, Celery, Apache JMeter, Mitm Proxy, Git, LaTeX, Wireshark

    Machine Learning Frameworks

    Tensorflow, Langchain, PyTorch, Keras, Pandas, Scikit-learn

    Project

    Youtube Transcript Summarizer using RAG

    Machine Learning & Web Project

    • Built a complete end-to-end application to summarize YouTube videos using transcripts with a Retrieval-Augmented Generation (RAG) pipeline.
    • Integrated a locally hosted LLM model via Ollama and constructed the RAG pipeline using LangChain to retrieve relevant transcript information and generate context-aware answers.
    • Developed APIs using Django to serve transcript summaries and user queries through backend service interface.
    Tools:
    Langchain, Django, Ollama
    October 2024

    E-Cloud-Learniverse

    Cloud-Native DevOps Project

    • Developed a comprehensive three-tier web application (FastAPI backend, React frontend and PostgreSQL database) demonstrating modern cloud-native development practices.
    • Containerized the entire application stack using Docker and Docker Compose, achieving consistent deployments and horizontal scalability across multiple environments.
    • Orchestrated production-ready deployment on Kubernetes using Helm charts, implementing industry-standard DevOps practices for container management and service discovery.
    Tools:
    Docker, Kubernetes, Helm, FastAPI, ReactJS, PostgreSQL
    Aug 2024

    Movie Recommender System

    Machine Learning & Web Project

    • It is a content-based machine learning recommendation model using a dataset of approximately 5000 movies.
    • The dataset was pre-processed and features were extracted using BoW (Bag of words) vectorization technique.
    • The frontend and API was developed in Django framework.
    Tools:
    Pandas, Sklearn, Django
    September 2021

    Redis Replication & High availability with Redis Sentinel

    Docker Project

    • It is a Docker configuration project to run 1 Master & Two Slave Replica instances within a Docker Network & for high availability of these Redis servers, 3 Redis Sentinel Docker instances are used within the same Docker Network.
    • Also, utilized HAProxy as a Layer-4 (Transport Layer) proxy to perform health checks on the master Redis node and consistently route external application clients' requests to the latest master Redis node.
    • Orchestrated the Redis cluster using Kubernetes StatefulSets and manifests, ensuring persistent identities, stable storage and efficient Pod scalability.
    Tools:
    Redis, Redis-Sentinel, Docker, HAProxy, Kubernetes
    January 2024

    Simple Custom Linux Read Driver

    C Project

    • Developed a custom Linux kernel driver to handle user-space read operations, returning a string acknowledgment from kernel space to user space. It is built to gain knowledge about how Linux kernel drivers work inside.
    Tools:
    C, Multipass (for virtualizing the OS)
    August 2024

    p2p Messenger

    Android Project

    • p2p Messenger is an instant messaging Android application that is built using Java socket and has peer‑to‑peer connection architecture in it.
    Tools:
    Android SDK, Java
    November 2019

    E-Learniverse

    Web Project

    • It is a self project to learn & document the learnings of Backend Application frameworks & tools.
    • Containerized the application with the help of Docker & Docker-compose.
    • Used Celery & Celery-Beat for doing asynchronous & periodic tasks in Django.
    • Steps & basic guide to learn & apply Flexbox for basic UI design.
    • Sending one-time, scheduled & bulk android notifications to corresponding demo app
    • Basic Kubernetes has been used to run & scale the web-application for deployment
    Tools:
    Django, Docker, Docker-compose, Celery, Redis, Rabbit-MQ, Kubernetes
    January 2022

    Programming

    Overall solved 600+ different problems
    • Codeforces: 350+ solves
      🔗
    • Leetcode: 310+ solves
      🔗
    • Vjudge: 250+ solves
      🔗
    • UVa Judge: 80+ solves
      🔗
    • StopStalk: akash_podder🔗
    • HackerRank: akashpodder🔗

    Certification

    • HackerRank Problem Solving (Basic) Certificate 🔗
    • HackerRank Software Engineer Certificate 🔗

    Award

    2nd Runner-Up in Software Hackathon held on 19-20th April at LICT SUST TECHFEST 2019

    • Developed a sign language conversion Android App called "Echo" that seamlessly translates between text, speech and American Sign Language, enabling bidirectional communication to help disabled people in the 48 hour Hackathon

    Scholarship

    University Merit Scholarship

    Shahjalal University of Science and Technology
    2017-2018