Summary
Overview
Work History
Education
Skills
Certification
Projects
Timeline
Generic

Srinivas Prashanth Kandula

Boise,ID

Summary

Innovative and results-driven Data Scientist and Machine Learning Engineer with extensive experience in developing and deploying scalable machine learning models and frameworks in hybrid cloud environments. Proficient in creating microservices, utilizing advanced data processing techniques, and implementing robust monitoring and performance tracking solutions. Demonstrated expertise in optimizing machine learning pipelines, enhancing model accuracy, and driving operational efficiency through automation and intelligent systems.Co-authored an IEEE publication on multi-modal deep learning, contributing to cutting-edge research and advancements in the field.

Overview

8
8
years of professional experience
1
1
Certification

Work History

Machine Learning Engineer

Micron Technology, Inc.
Boise, ID
06.2022 - Current
  • Developed microservices as part of a generic framework for a model-agnostic machine learning inference system, facilitating deployment of ML models to hybrid cloud environments such as Red Hat OpenShift and GCP using Kubernetes, Docker containerization, and Cookiecutter for project scaffolding.
  • Engineered an event-driven machine learning inference framework that communicates with Solace or GCP Pub/Sub, utilizing KEDA for scalable microservices in hybrid cloud environments.
  • Developed a Splunk-based dashboard to monitor the performance of the deployed framework, providing insights into Solace message metrics, model inference logs, and pod-related performance metrics for enhanced operational visibility
  • Utilized Apache Arrow and Flight Server to extract data via command line tools, exposing the data extraction process as a microservice for enhanced accessibility and efficiency in data operations.
  • Engineered ETL processes to extract sensor data from on-premises systems and ingest it into BigQuery on GCP, supporting data analysis and reporting, while leveraging Docker and GKE for efficient container management.
  • Engineered ETL processes to extract sensor data from on-premises systems and ingest it into BigQuery on GCP, supporting data analysis and reporting, while leveraging Docker and GKE for efficient container management.
  • Designed and implemented custom metric collection strategies using Stackdriver Monitoring to track key performanceindicators (KPIs) specific to application and infrastructure health.

Senior Software Engineer-ML

Infrrd Pvt Ltd
Bengaluru, India
04.2020 - 08.2021
  • Engineered an Intelligent Document Processing (IDP) platform to automate invoice text extraction, achieving a 85% accuracy rate and reducing manual effort by 30%, thus streamlining operational processes.
  • Engineered a solution to index invoice and annotation data using Elasticsearch, preventing redundancy and facilitating optimized data access for image recognition model training and inference.
  • Implemented MLflow to manage model versioning throughout the training and inference phases, enhancing traceability and reproducibility of machine learning models.
  • Executed row and column segmentation on tabular data utilizing Fully Convolutional Networks (FCNs), improving the precision of data extraction for analysis and model training.

Data Scientist

Virtusa Consulting Services Private
Bengaluru, India
03.2019 - 04.2020
  • Analyzed British Telecom's operational processes to identify opportunities for automating workflows, leveraging machine learning-based prediction models to enhance efficiency and streamline operations.
  • Developed a classification model to predict modification types for customer orders, transforming relevant data and achieving a 20% increase in prediction accuracy, leading to a 15% reduction in service-level determination time.
  • Collaborated on deploying machine learning models in a Java application via Docker, ensuring efficient containerization and streamlined integration into the existing architecture.
  • Utilized SHAP and LIME to interpret machine learning model predictions, enhancing business understanding and guiding data-driven decision-making through detailed insights.

Data Scientist

Wipro Technologies
Bengaluru, Karnataka
06.2018 - 02.2019
  • Developed and optimized an image recognition model, resulting in a 30% increase in Intersection over Union (IOU), thereby significantly improving the accuracy and performance of object detection tasks. Utilized advanced techniques such as convolutional neural networks (CNNs), data augmentation and hyperparameter finetuning to achieve this enhancement.
  • Performed image preprocessing on anomaly detection videos using OpenCV, employing techniques such as motion vectors and various edge detection algorithms to prepare high-quality training data, thereby enhancing model accuracy and robustness.
  • Trained a deep learning model for text extraction, specifically designed to accurately extract length information from each image, improving data extraction efficiency and accuracy for downstream processing tasks.
  • Developed and implemented a data augmentation technique to enhance model robustness and performance in scenarios with scarce training data.

Data Scientist

Wipro Technologies
Hyderabad, India
01.2018 - 05.2018
  • Conducted feature extraction to identify key predictors of customer loan defaults for a leading banking services company, resulting in a model that achieved an accuracy of 85% and a F1 score of 0.78, significantly enhancing risk assessment and decision-making processes.
  • Executed A/B testing to compare loan default rates pre- and post-implementation of the prediction model, achieving a 15% reduction in defaults and validating the model's effectiveness in real-world scenarios.
  • Implemented SMOTE to address class imbalance, generating synthetic samples that improved model recall by 15%.
  • Conducted hyperparameter tuning using Grid Search and Random Search, resulting in a 10% improvement in accuracy.

Project Engineer

Wipro Technologies
12.2016 - 12.2018
  • Facilitated end-to-end project support for the business, managing all phases from initial discussions and requirements gathering to design, coding, testing, debugging, implementation, and integration, aligning solutions with strategic goals.
  • Conducted thorough code debugging on Microsoft CRM plugins and workflows, along with SQL Server stored procedures, identifying and resolving issues to maintain system integrity and functionality.
  • Developed detailed implementation plans for change management, including timelines, key milestones, and resource requirements to facilitate smooth transitions and minimize disruptions.

Education

MASTERS IN DATA SCIENCE AND ANALYTICS - Computer Science And Programming

GEORGIA STATE UNIVERSITY
Atlanta, GA
12-2022

Skills

  • Google Cloud Platform
  • PySpark
  • Kubernetes
  • Data Science
  • Computer Vision
  • NLP
  • Machine Learning
  • C#
  • Elastic search
  • Python
  • OpenCV
  • Data visualization
  • SQL query language
  • R
  • Inferential Statistics
  • Hypothesis testing
  • Multivariate Data Analysis
  • Dimensionality Reduction

Certification

  • Multi-modal deep learning on imaging genetics for schizophrenia classification- IEEE, ICASSP’23
  • Deep Learning on Multi-modal Neuroimaging Data for Schizophrenia Classification- OHBM’23

Projects

  • Implemented Deep Learning based Movie Recommender System on Movie Lens Dataset using Graph Convolution Network
  • Implemented predictors for classification approach using Tree based ensemble algorithms to estimate the worthiness of granting personal loan on German banking dataset and Ranking features based on Feature Importance scores.

Timeline

Machine Learning Engineer

Micron Technology, Inc.
06.2022 - Current

Senior Software Engineer-ML

Infrrd Pvt Ltd
04.2020 - 08.2021

Data Scientist

Virtusa Consulting Services Private
03.2019 - 04.2020

Data Scientist

Wipro Technologies
06.2018 - 02.2019

Data Scientist

Wipro Technologies
01.2018 - 05.2018

Project Engineer

Wipro Technologies
12.2016 - 12.2018

MASTERS IN DATA SCIENCE AND ANALYTICS - Computer Science And Programming

GEORGIA STATE UNIVERSITY
Srinivas Prashanth Kandula