Machine Learning Software Engineer

CHAITANYA UPPULURI

+1-716-***-**** Hackensack, NJ

Email me Linkedin Github

OBJECTIVE

Software Engineer with 3+ years of experience in software, developing and maintaining front end with HTML, CSS, JS, Node, and implementing security measures through AWS IAM roles, OSSEC and python, and have published a research paper on LSTM Networks for Sentiment Analysis in Smart DSC 2019 . EDUCATION

Bachelors of Computer Science, Vignan’s University 2015-2019 GPA: 9.5

Masters of Computer Science, University at Buffalo 2021 - 2022 SKILLS

Technical Skills: Python, R

Frameworks and Libraries: Django, ReactJS, NodeJS, Bootstrap Frameworks and Libraries: Machine Learning: TensorFlow, PyTorch, Scikit-learn, Keras NLP: Hugging Face Transformers, SpaCy, NLTK

Computer Vision: OpenCV

Big Data: Apache Spark, Hadoop, Hive

Cloud Platforms: AWS SageMaker, Google AI Platform, Azure ML Tools and Platforms: Data Visualization: Tableau, Matplotlib, Seaborn MLOps: MLflow, Jenkins

Containerization: Docker, Kubernetes

Version Control: Git, GitHub/GitLab

Pre-trained Models: GPT, BERT, T5, RoBERTa, ResNet, EfficientNet Cloud Services: AWS Lambda, CloudWatch, S3

Tools: VSCode, Jupyter Notebook, Replit, APIs

Security OSSEC, Python

Other: Object-Oriented Programming, Data Structures, Algorithms, A EXPERIENCE

Full Stack AI/ML Engineer Mar 2023 - Current

Agile Squad Inc. Hackensack, NJ

• Designed, developed, and deploy machine learning models for various applications (e.g., predictive analytics, NLP, computer vision).

• Built and maintained ML pipelines for data ingestion, preprocessing, feature engineering, model training, eval- uation, and deployment.

• Optimized models for performance, scalability, and real-time inference.

• Collected, clean, and preprocess large datasets from diverse sources (structured and unstructured).

• Performed exploratory data analysis (EDA) to identify trends, patterns, and insights.

• Developed NLP models for tasks like sentiment analysis, text summarization, and machine translation.

• Implemented CI/CD pipelines for automated model training, testing, and deployment.

• Monitored model performance in production and retrain models as needed.

• Used containerization tools (e.g., Docker, Kubernetes) for scalable deployment.

• Collaborate with data engineers to ensure data quality and availability. Software Development Engineer Jul 2022 - Feb 2023

Amazon.com Santa Monica, CA

• I worked in team ADAM-SHIELD, a child team in Amazon Ads responsible for managing authentication and permissions in Amazon ads. Amazon ads provide and maintain consoles and accounts to different users for publishing product and brand ads to Amazon sites. The team manages the permissions for those accounts.

• My job role was focused in Talos, a hierarchical system built from Amazon’s Cloud Directory Service, to manage permissions for different user accounts. I worked on migrating to a new version of Talos, providing an easier way to manage, authorize and access stored user permissions. My role included tasks such as creating lambdas for older scripts during migration using Java, JavaScript, and TypeScript, creating metrics, alarms and dashboards using Amazon Web Services for detecting, monitoring and analyzing anomalies and service health, and addressing problems encountered by internal customers with other services owned by our team as part of on-call process. PERSONAL PROJECTS

Web Development

• Industrial waste material management (BACHELORS): Many industries and manufacturers produce products which are unnecessary or leftovers in the end. Any company in need of such byproducts can reach out to a company with abundant stock rather than manufacturing. Our website acts as a medium for such inter- industrial trade with a consumer-supplier interaction. Implemented the system using Django for backend logic and React for an interactive front-end interface, improving material trade efficiency by 40 percent.

• Corona Impact On India: The website provides details regarding the COVID cases in India which include

(Active, Deceased, Recovered cases). It also gives each state COVID status in India and state-wise ranking in case count.

Databases

• Airborne Wildlife conservation from aircraft strike Database (MASTERS): Aircrafts often report bird strike accidents. So the database provides data regarding the airline routes, airline timings which suffered accidents. The database shows which kind of birds took more hits, how extinct they are. This data can be useful for Airborne wildlife conservation organizations, research departments of airlines to reduce aircraft damage or accidents.

• Database for Industrial Waste management (BACHELORS): The database contains data on purchase history of products, surplus, company offering and companies bought or willing to buy. Cloud Computing

• Open project with Amazon Web Services: Installing Apache server and mySQL database in an AWS infrastruc- ture service. Uploaded corpus data into it through shell. Achieved robust security and performance improve- ments using IAM roles and VPCs for network isolation, reducing data processing time by 25 percent.

• Hadoop Big Data and Tkinter (MASTERS): Obtaining author and co-authors of a given book - Given a book in the databases from different nodes, the names of author and co-authors are fetched using this project. A Tkinter application is developed to fetch the author and co-authors of a book. Machine Learning

• Sentiment Analysis for Sarcastic Reviews Using LSTM Networks: Sentiment analysis can identify good and bad reviews, but not sarcastic reviews. The project provides a solution to identifying such reviews. The objective was to increase the probability of correct prediction of sarcastic reviews with optimum accuracy rate beyond 80 percent ef£ıciency

• Panorama creation project (Image Stitching): The project is done using openCV Given images, the openCV model can align images according to detecting common link points between images and attach them side-by-side, making it a £ınal panorama image. Optimized the algorithm for accuracy and performance, reducing processing time by 20 percent.

• Anomaly detection in Border Gateway Protocol: The Border Gateway Protocol is the most widely used network protocol. Its suffered three most deadly anomalies in its past namely, Moscow blackout, Wannacrypt and Slammer. The project is about detecting the anomaly (if it exists), given all the router and pathway information of a BGP route. The model analyzed raw network traffic, detecting anomalies with a 90 percent accuracy using SVM and random forest classifiers. Built using Python, Pandas, and Scikit-learn for data processing and model training

• Credit card fraud detection: Developed a predictive model to assess the risk of credit card fraud based on transaction data. Leveraged ensemble learning techniques, achieving 85 percent accuracy in predicting fraudulent activities, and improved fraud detection rates by 20 percent. Technologies used: Python, Scikit-learn, and Pandas for data analysis and model building.

• MNIST number prediction: Built a neural network using TensorFlow to recognize handwritten digits from the MNIST dataset. The model achieved 98 percent accuracy by training on large datasets and optimizing hyperparameters. This project demonstrated proficiency in deep learning techniques for image classification.

• Cancer disease prediction: Developed a machine learning model that predicts the likelihood of cancer develop- ment based on patient health data. The model achieved over 85 percent accuracy, enabling early-stage cancer detection through classification algorithms. The project used Python, Scikit-learn, and Matplotlib for data processing, visualization, and prediction.

CERTIFICATIONS

Full Stack Development

• Front End Development - HTML (Great Learning)

• Front End Development - CSS (Great Learning)

• How to Build your own Chatbot using Python? (Great Learning) Cloud computing

• Oracle Cloud Infrastructure Foundations 2020 (certified associate)

• Cloud Computing with Amazon Web Services(Great Learning)

• Cloud Foundations (Great Learning)

Machine Learning

• Machine Learning Foundations (Great Learning)

• Machine Learning by Andrew NG (Coursera)

• ChatGPT for Coders (Great Learning)

Machine Learning

• Practical Intrusion Detection using OSSEC (EC-Council)

• Cybersecurity information protection (INFOSEC)

• Python for Cybersecurity (INFOSEC)

• GO Programming Language (Great Learning)

Contact this candidate