Tarun Kumar Pothineni profile picture

Hello, I'm

Tarun Kumar Pothineni

Full Stack Developer, Java Developer

My LinkedIn profile My Github profile
Arrow icon

Get To Know More

About Me

Profile picture
Experience icon

Experience

2.5+ years
Full Stack Engineer

Education icon

Education

Master's in CS - State University of New York at Binghamton
Bachelor's in ECE - LNMIIT

Full-stack developer with 2.5 years of experience in microservices architecture, database management, and Agile method- ologies. Pursuing an MS in Computer Science at the State University of New York, Binghamton, with a proven ability to optimize system performance, enhance AI models, and drive business impact.

Arrow icon Arrow icon

Explore My

Skills

Programming Languages, Testing, Databases and Methodologies

Experience icon

Java

Experience icon

Python

Experience icon

C

Experience icon

C++

Experience icon

HTML

Experience icon

CSS

Experience icon

R Programming

Experience icon

MySQL

Experience icon

PostgresSQL

Experience icon

Oracle

Experience icon

VoltDB

Experience icon

SDLC

Experience icon

Agile

Experience icon

Waterfall

Cloud Technologies, Version Control Tools, CI/CD and DevOps Tools

Experience icon

AWS

Experience icon

EC2

Experience icon

Git

Experience icon

Github

Experience icon

Jenkins

Experience icon

Docker

Experience icon

Kubernetes

Experience icon

Bitbucket

Experience icon

Kafka

Experience icon

Grafana

Experience icon

Kibana

Experience icon

Terraform

Experience icon

Prometheus

Frameworks and Developer Tools

Experience icon

Spring Boot

Experience icon

Spring

Experience icon

Hibernate

Experience icon

React

Experience icon

Angular

Experience icon

Maven

Experience icon

RESTful API

Experience icon

gRPC

Experience icon

V S Code

Experience icon

Eclipse

Experience icon

Intellij

Experience icon

Postman

Experience icon

DBeaver

Experience icon

Office 365

Arrow icon Arrow icon

Explore My

Experience

Senior Engineer

Comviva Technologies Limited. (Dec 2022 - Jul 2023)

  • Spearheaded a Proof of Concept to upgrade the database, slashing costs by 66% and boosting efficiency by 15% compared to VOLTDB, while migrating 23 microservices seamlessly to Singlestore.
  • Designed and launched a Purchase History-based Offer Holdout feature, driving a 17% increase in client revenue by optimizing offer distribution to recent customers.
  • Developed a Lambda architecture for real-time offer delivery, accelerating traffic by 60% and amplifying revenue by 40%, while creating a KTable and KStreams accumulator to expedite transaction processing and offer allocation

Engineer

Comviva Technologies Limited. (Jul 2021 - Nov 2022)

  • Revamped an existing microservice by integrating caching, streamlining database queries, and optimizing code, reducing customer offer delivery time by 71.4%.
  • Integrated all microservices with Consul, removing reliance on external configuration files and enhancing multi-deployment compatibility and seamless inter-service communication.
  • Deployed the EFK stack for centralized logging , improving log collection, aggregation, and analysis by 40%.
  • Configured and launched Grafana for real-time monitoring and visualization,increasing system observability by 30%.
  • Optimized PostgreSQL queries in a microservice, boosting efficiency and reducing output response time by 600ms.

Engineer Intern

Comviva Technologies Limited. (Jan 2021 - Jun 2021)

  • Engineered a microservices architecture that facilitated the delivered personalized offers to customers via REST API; improved response times by 40% while handling requests from 15,000+ active users daily.
  • Refined code quality and reduced production defects by achieving 80% code coverage through test-driven development.
  • Resolved over 300+ real-time bugs during client-server deployment, enhancing system stability and performance.
Arrow icon Arrow icon

Browse My Recent

Projects

HandWritten Digits Recognition

Python ,CNN ,TensorFlow Lite ,Java ,Android Studio

  • Developed an Android app using Java and TensorFlow Lite to predict handwritten digits with a custom CNN model, achieving 98.97% accuracy and reducing latency to 2ms.
  • Evaluated inbuilt models such as ResNet50, and MobileNetV3, achieving accuracies of 98.55% and 96.64%, respectively.
  • Implemented a robust machine learning solution with a 4-layer CNN, trained over 7 epochs using the Adam optimizer and dropout regularization, outperforming traditional models.

Enhanced YOLO Framework

Python ,PyTorch ,CUDA ,NumPy ,Pandas , Matplotlib ,Google Colab

  • Optimized the YOLOv7 model, improving mean Average Precision (mAP) by 3.1% and reducing inference time by 5.4%, enhancing detection speed and accuracy on the Pascal VOC 2012 dataset.
  • Increased model precision by 2.4% and recall by 1.8% through hyperparameter tuning and efficient data preprocessing.
  • Reduced GPU memory usage by 4.5% and enhanced overall model efficiency, accelerating training cycles on Google Colab with a T4 GPU.
Arrow icon Arrow icon

Get in Touch

Contact Me

Arrow icon