1. Herambeshwar Pendyala
hpend001@odu.edu | linkedin.com/in/Heramb-Pendyala | github.com/Heramb001 | 732-476-9255
Computer Science Graduate student. Actively looking for internship opportunities in an organization where I can apply data science skills to
solve complex problems and develop business models driven by technology. 2 years of experience working in Natural Language Processing,
big data technology, agile software development environment.
EDUCATION
Master of Science - Computer Science - Old Dominion University (ODU) (GPA:3.92/4.00) (Jan 2019 – Dec 2020)
Bachelor of Technology - ECE - Lovely Professional University (LPU) (GPA:3.42/4.00) (Aug 2012 – May 2016)
TECHNICAL SKILLS
Programming Languages : Java, Python, C, C++, UNIX Shell Script
Big Data Ecosystem : Spark, Hadoop, MapReduce, HDFS
Frameworks & Libraries : Numpy, Pandas, Matplotlib, Scipy, NLTK, Scikit-learn, Tensorflow, Keras, Django, OpenCV
Web technologies : HTML5, CSS3, Bootstrap, JavaScript, JSON, XML, PHP, Node JS, Express
Databases : Oracle 12c, SQL, MySQL, Mongodb
Tools & IDEs : Eclipse, Microsoft Visual Studio code, WinSCP, PuTTY, Postman, SoapUI, Oracle SQL Developer, Toad, GIT
PROFESSIONAL EXPERIENCE
Graduate Research Assistant, Information Technology Services, Old Dominion University (May 2019 – Present)
● Diva - Chatbot | Virtual Assistant: Developed FAQ chatbot for Instruction Technologies course, to provide students with automated
responses to course related queries such as assignment deadlines, module details; augmented with a closed loop feedback system to improve
chatbot performance. Technologies: Google DialogFlow, Node JS, MySQL.
Graduate Research Assistant, Old Dominion University (May 2019 - Aug 2019)
● Parallel and distributed computing using raspberry pi cluster: Built a 4 node Raspberry Pi cluster to analyze parallel and distributed
computing. Implemented different algorithms on multiple nodes to reduce execution time by 50% in applications such as face detection,
Homographic encryption on images, Log mining. Technologies: mpi4py, opencv, slurm.
Graduate Teaching Assistant, Old Dominion University (Jan 2019 – May 2019)
● Created a virtual Snakes and Ladders board game applying object oriented concepts in C++ taught in the course; TA duties also included
developing engaging and challenging programming quizzes and assignments. Technologies: C++
Software Engineer – Tech Mahindra, Noida, India. (Jul 2016 – Dec 2018)
Successfully completed several Proof of Concepts leveraging Machine learning techniques to speed up deliverables, including:
● Skill Sphere | Data Analytics: Using information retrieval tools, improved performance of model that ranked users based on technical
skills parsed from resumes, and created a supporting webservice. Technologies: SOLR, Kibana, ElasticSearch, Python, Django.
● Customer Query Clustering | Data Science: Built a k-means clustering model with 95% accuracy to categorize user queries for feature
engineering on text data using Natural Language Processing techniques like TF-IDF. Technologies: python, NLTK, Scikit Learn
● EDGE - Dispatch and Scheduling Engine | Java: Automated webservices and microservices testing processes using Maven based
framework. Web scraping using selenium to perform UI validations. Technologies: Java, Maven, Unix Shell Scripting, Postman, SOAPUI.
GRADUATE EXPERIENCE
Learning Analytics - Predict Students at risk | Data Mining: Achieved 85% test accuracy in predicting student course failure rate,
evaluated model performance using ROC curve. Built machine learning model such as Random forest, Decision Tree, naive bayes classifier
through Exploratory data analysis, data cleaning and data preprocessing. Technologies: Python, Scikit-Learn, Weka.
Automatic Hand Sign Detection | Deep Learning: Applied techniques of hyperparameter tuning, regularization, adam optimization to
build a sign language recognizer using CNN (Convolution Neural Network) model from scratch, achieving 85% accuracy on a test set.
Technologies: Python, Tensorflow
Happy House Challenge | Deep Learning : Built a CNN (Convolution Neural Network) model to detect happy faces in images using
adam optimizer, achieving 95% test accuracy (and 99% train accuracy). Technologies: Python, Keras
Chatbot | Deep Learning : Developed a chatbot with Natural Language Processing techniques by building a Seq2Seq model using RNN
(Recurrent Neural Network) on a Cornell movie database to generates phrases which can be used as replies. Technologies: Python, tensorflow.
ACHIEVEMENTS
Runner up LPU Project Expo, 2016, Project AVI (Assistant for Visually impaired) - developed an IOT device that uses data collected from
sensors to give feedback to visually challenged users about their surroundings.