Software Engineer

Jayath Kumar Lekkala

747-***-**** ********@*******.***

EDUCATION

Stevens Institute of Technology, Hoboken, NJ May 2021 Master of Science in Computer Science with Specialization in Data Science, GPA: 3.44/4 Gitam University, Visakhapatnam, India April 2014

Bachelor of Technology in Computer Science, GPA: 7.49/10 TECHNICAL SKILLS

Programming Languages: Python, Java, SQL Database Systems: MySQL Machine Learning: Pandas, Scikit-learn, NumPy Testing Tools: Selenium, Appium Web Programming: HTML, CSS, JavaScript Natural Language Processing: NLTK, SpaCy Visualization: Tableau, Seaborn, Matplotlib

Deep Learning: Keras, TensorFlow

Statistical Packages: Pandas, Numpy, SciPy

PROFESSIONAL EXPERIENCE

EBUTOR Distribution Private Limited, Hyderabad, India – Software Engineer June 2014 – August 2017

• Deployed a recommendation system to production for an e-commerce website to estimate and predict retailer purchase behavior that led to a rise in revenue by 2-fold Y-o-Y between 2015 to 2017

• Implemented SQL queries dynamically on the database and designed a model in Python to build a recommender system using machine learning that aids in recommending products to end retailers

• Addressed a growing problem of carding on the e-commerce store and solved the problem by building a fraud detection system using classification algorithms that aided in reducing fraud transactions by 50%

• Collaborated with the managers to assist the product and management teams in onboarding more customers by bringing in advanced technologies that resulted in a rise of its sales by 20% for the next 2 years

• Mentored a team to incorporate the recommendation system created earlier into the mobile application through an API leading to an increase in number of application downloads by 10,000 over a quarter ACADEMIC PROJECTS

Multimodal Hateful Meme recognition

• Developed an algorithm using transfer learning that recognizes hateful content in memes and achieved an accuracy of 64% by executing a multi-modal approach using Convolutional neural network and BERT

• Formulated a process to extract textual data into machine encoded text from a subtitle text superimposed on an image using Optical Character Recognition

• Executed several preprocessing techniques to clean images using OpenCV and Pillow as well as textual data used for model input

ASHRAE Great Energy Predictor Kaggle Competition

• Implemented a time series forecasting technique to forecast the energy consumption on a metered building using an Autoregressive Moving Average model and linear regression model with an accuracy of 82%

• Lead a Team of 3 for a course project to implement different methodologies on a time series dataset that can identify useful insights for analyzing the given problem

• Presented a detailed analysis of our team’s findings towards the end of the coursework Sentiment Analysis for analyzing Disaster Tweets

• Created an algorithm to classify whether a tweet is about a disaster using Natural Language Processing and accomplished an accuracy of 77%

• Analyzed the tweets using exploratory data analysis and data visualization to discover insights in the data

• Predicted the tweet sentiments using classification and word2vec modelling

Contact this candidate