Sahana Alliyandiru Jayasheela
Summary
Highly skilled machine learning engineer and data scientist with over 7.5 years of experience in implementing, designing, and deploying AI-driven solutions. Proficient in Gen AI and LLM technologies, including deploying models like BERT, GPT, and AutoML pipelines for real-world applications. Expertise in MLOps workflows for scalable AI systems, leveraging cutting-edge technologies such as TensorFlow, PyTorch, Kubernetes, and Docker. Deep experience with NLP techniques, predictive modeling, deep learning algorithms, and big data tools like Apache Spark, Hadoop, and Kafka. Proven track record of delivering impactful business outcomes in finance, logistics, healthcare, and telecommunications industries. Demonstrates exceptional problem-solving and collaboration skills with a passion for data analysis, machine learning, and deploying intelligent systems. Adept at driving projects from concept to execution, optimizing pipelines, and developing actionable insights for critical decision-making.
Overview
10
10
years of professional experience
1
1
Certification
Work History
Sr Data Scientist
United Parcel Service
01.2023 - 12.2024
- Led the development and deployment of advanced AI and machine learning solutions for various applications
- Designed scalable data pipelines on Google Cloud Platform (GCP) and containerized models using GKE and Docker
- Executed and validated ETL workflows with Dataflow and BigQuery, ensuring data accuracy and optimizing processes for scalability
- Spearheaded the creation of ACES (Address Classification Enrichment Service) using pre-trained BERT models, leveraging AutoML techniques to optimize model performance and accuracy
- Developed ADAPT US/INTL (Anticipated Delivery and Package Time) and HAPI (Harmonized Automated Predictive Intelligence) frameworks using TensorFlow and PyTorch, integrating MLOps workflows to automate deployment
- Built HARTS (Harmonized AI Respondent Tariff System), an ML-based system for predicting product classification codes, achieving a 25% improvement in accuracy
- Created a GEN AI demo project using Terraform for infrastructure management and cloud deployment
- Utilized Bayesian inference for real-time decision-making in A/B testing, improving engagement rates by 20%
- Implemented CUDA-accelerated training pipelines, reducing training time by 50%
- Developed dashboards in Looker to track application performance and business KPIs
- Integrated Dataiku with GCP to streamline model development and deployment, enhancing collaboration and scalability.
Software Engineer
AT&T
06.2022 - 01.2023
- Designed and deployed ontology-based recommendation systems, enhancing semantic search capabilities for knowledge graphs
- Built and optimized ML pipelines in Azure Databricks, achieving a 40% improvement in deployment efficiency
- Migrated ERP datasets to Azure Synapse using Azure Data Factory, enabling advanced analytics and reporting
- Devised a semantic search system to recommend subscription plans based on customer review insights, leveraging NLP and graph traversal techniques.
Data Scientist
Qdata Inc
01.2022 - 06.2022
- Developed predictive analytics models for fraud detection, using Logistic Regression and Random Forest to improve anomaly detection rates by 30%
- Built Grafana dashboards to monitor ETL workflows and cloud infrastructure performance in real-time
- Integrated Kubeflow pipelines with GCP Vertex AI for end-to-end machine learning model deployment.
Machine Learning Engineer
igeeks
12.2021 - 01.2022
- Developed machine learning models using techniques like clustering analysis, market analysis, association rules, and recommendation systems
- Built predictive models such as Decision Trees, Logistic Regression, and Support Vector Machines to enhance business insights
- Collaborated with data science teams to prioritize tasks and provide acceptance criteria
- Designed and implemented analytics pipelines for graph-based data models, improving data processing efficiency
- Optimized data analytics workflows using advanced machine learning algorithms, enabling actionable insights for stakeholders
- Enhanced model accuracy and scalability by integrating automated workflows and deploying models to production environments.
Machine Learning Engineer
Bank of America
11.2020 - 11.2021
- Collaborated with the data science team to design and build analytics models for fraud detection and financial risk assessment
- Developed and deployed machine learning models, including Logistic Regression, Random Forest, and XGBoost, to identify fraudulent transactions with 95% precision
- Created graph schemas using GSQL and implemented algorithms such as PageRank and Louvain to uncover fraud rings, enhancing the fraud detection process
- Streamlined data extraction, transformation, and loading (ETL) processes using AWS Glue and Amazon Redshift, reducing manual effort by 40%
- Automated data validation and reporting tasks with AWS Lambda, ensuring data accuracy and integrity
- Optimized ETL performance and scalability using Amazon CloudWatch for real-time monitoring and resource management
- Integrated TensorRT into deep learning pipelines, achieving a threefold reduction in inference latency for real-time applications
- Deployed TensorRT models to edge devices for enhanced scalability and responsiveness
- Used AWS S3 for secure data storage and employed Glue scripts for efficient data processing
- Improved business operations by implementing advanced statistical methods like A/B testing and anomaly detection to generate actionable insights
- Developed and maintained dashboards in Tableau to track model performance and provide stakeholders with clear visualizations of key metrics.
Machine Learning Engineer
Avaya Communications
02.2020 - 11.2020
- Designed ARIMA and XGBoost-based models for sales forecasting, improving forecast accuracy by 24%
- Developed distributed ML models for recommendation systems, leveraging Kubernetes for cloud-based deployment
- Trained deep neural networks for customer segmentation, using GPU-accelerated TensorFlow pipelines to enhance computational efficiency.
Machine Learning Engineer
Tuutkia LLC
12.2019 - 02.2020
- Implemented a graph-based recommendation system using Neo4j, enabling personalized product suggestions
- Created risk assessment models using Bayesian inference to identify key drivers of business performance.
Research Assistant
SJSU
01.2019 - 05.2019
- Designed real-time pothole detection models using CNN and RNN architectures, improving classification accuracy
- Built data pipelines for large-scale datasets, enabling efficient preprocessing and feature engineering workflows.
Associate Software Engineer
Mphasis Ltd
01.2017 - 05.2017
- Built web-based applications for the banking domain using HTML, CSS, and JavaScript
- Developed recommendation systems using Neo4j and automated data workflows with MySQL.
Data Analyst
Pavan Informatics
03.2015 - 01.2017
- Led the development and deployment of advanced AI and machine learning solutions for various applications
- Designed scalable data pipelines on Google Cloud Platform (GCP) and containerized models using GKE and Docker
- Executed and validated ETL workflows with Dataflow and BigQuery, ensuring data accuracy and optimizing processes for scalability
- Spearheaded the creation of ACES (Address Classification Enrichment Service) using pre-trained BERT models, leveraging AutoML techniques to optimize model performance and accuracy
- Developed ADAPT US/INTL (Anticipated Delivery and Package Time) and HAPI (Harmonized Automated Predictive Intelligence) frameworks using TensorFlow and PyTorch, integrating MLOps workflows to automate deployment
- Built HARTS (Harmonized AI Respondent Tariff System), an ML-based system for predicting product classification codes, achieving a 25% improvement in accuracy
- Created a GEN AI demo project using Terraform for infrastructure management and cloud deployment
- Utilized Bayesian inference for real-time decision-making in A/B testing, improving engagement rates by 20%
- Implemented CUDA-accelerated training pipelines, reducing training time by 50%
- Developed dashboards in Looker to track application performance and business KPIs
- Integrated Dataiku with GCP to streamline model development and deployment, enhancing collaboration and scalability.
Education
Deep Learning, Machine Learning, Big Data Engineering, Large Scale Analytics, Data Mining -
MS - Computer Engineering (Data Science
San Jose State University
San Jose, CA2019
BS - Electronics & Communication Engineering
Vidya Vardhaka College of Engineering
Mysore2016
Skills
- Technical Skills
- Machine Learning: Logistic Regression, Support Vector Machines (SVM), Decision Trees, Random Forest, Gradient Boosting, XGBoost, K-Means Clustering, PCA, LDA, CNN, RNN, LSTM, GANs
- Natural Language Processing (NLP): Tokenization, TF-IDF, Named Entity Recognition (NER), Sentiment Analysis, Pre-trained Models (BERT, GPT, RoBERTa), Word2Vec, Hugging Face Transformers
- Programming: Python, R, SQL, Nodejs, JavaScript, C
- Cloud Platforms: AWS, GCP, Azure, Kubernetes, Docker
- Big Data & Tools: Apache Spark, Hadoop, Hive, Kafka, BigQuery
Certification
Certified TensorFlow Developer
AWS Certified Machine Learning Specialist
Google Professional Data Engineer
Timeline
Sr Data Scientist
United Parcel Service
01.2023 - 12.2024
Software Engineer
AT&T
06.2022 - 01.2023
Data Scientist
Qdata Inc
01.2022 - 06.2022
Machine Learning Engineer
igeeks
12.2021 - 01.2022
Machine Learning Engineer
Bank of America
11.2020 - 11.2021
Machine Learning Engineer
Avaya Communications
02.2020 - 11.2020
Machine Learning Engineer
Tuutkia LLC
12.2019 - 02.2020
Research Assistant
SJSU
01.2019 - 05.2019
Associate Software Engineer
Mphasis Ltd
01.2017 - 05.2017
Data Analyst
Pavan Informatics
03.2015 - 01.2017
Deep Learning, Machine Learning, Big Data Engineering, Large Scale Analytics, Data Mining -
MS - Computer Engineering (Data Science
San Jose State University
BS - Electronics & Communication Engineering
Vidya Vardhaka College of Engineering