My world transformed when I discovered data science; it was like unlocking a hidden superpower. With over 2 years of learning and a decade in SAP ERP, I've learned to see data as more than numbers and graphs—it's a canvas where stories of innovation and progress unfold. As an emerging Data Scientist, I thrive in unraveling complex data, exploring innovative solutions, and visualizing impactful insights.
My journey began with a simple curiosity about how the dormant data in servers and clouds could drive smarter decisions and change lives. This path led me to UNC Charlotte which not only honed me in developing models; I listen to the whispers of data, translating its wisdom into solutions that drive meaningful change.
Contact SakshiI’m on a mission to make a name for myself as a supportive and knowledgeable woman in data science.
My path to problem-solving in this field is characterized by a trio of key strengths : versatility, holistic thinker, and perseverance. Every day is a fresh opportunity to expand my horizons and apply a thorough and thoughtful approach to every problem I encounter.
Through my educational journey & internships, I've cultivated a solid foundation in the various stages of Machine Learning — from questioning and research to the intricate processes of data verification, cleaning, analysis, modeling, and performance assessment . These skills are not just theoretical; they're practical, demonstrated through the projects I've brought to fruition.
But my journey doesn't end with personal growth. I am equally passionate about contributing to the collective knowledge of the data science community. Through writing technical articles, I share the insights of my journey, fostering an environment where collaboration and knowledge-sharing are the cornerstones for collective advancement.
This is a Deep Learning for Meme Generation using PyTorch. It incorporated pre-trained models like ResNet50 and InceptionV3 for image encoding and GloVe embeddings for text. I improved text data with custom pre-processing methods and employed Beam Search with Top-k sampling for Meme Generation. The model's performance was evaluated using BLEU and BERT scores to gauge meme quality and relevance.
The primary aim of this project was to analyze the viewpoints of Elon Musk, the world's wealthiest individual (holding the top position at the time, as per Forbes' Real-Time Billionaires List), on a range of subjects and gauge public reactions to his statements. To achieve this objective, a dataset was sourced from Kaggle, comprising tweets from January 2022 to October 2022. The project encompassed various tasks, including emotion classification, topic extraction, and the identification of nouns and adjectives within the text data.
This project is a data visualization effort focused on Airbnb listings in Nashville, TN. We analyzed occupancy, pricing, seasonality, and ratings. The dashboard has two levels: the main page offers an overview of listing stats, while clicking on a listing provides details about the availability and reviews of that listing.
This project revolved around tackling a Kaggle competition called "Amex - Default Prediction," which had already concluded. Our primary goal was to achieve a top-tier score using a reduced set of features. To accomplish this, we employed various feature engineering techniques, including VIF (Variance Inflation Factor), PCA (Principal Component Analysis), and Shapley values. As a result of these efforts, we successfully achieved a score of approximately 0.77, coming close to the competition's best score of 0.80. Furthermore, we managed to significantly reduce the feature dimension from the original 190 to a more manageable 58, demonstrating the power of feature selection and engineering in improving predictive models.
This project involved data analysis for a well-known non-profit organization. We leveraged their social media data to conduct a comprehensive analysis of audience engagement, which enabled us to offer valuable recommendations for enhancing engagement strategies. Furthermore, we conducted a survey and utilized Charlotte's census data to identify distinct clusters based on individuals' propensity to volunteer and donate to this non-profit. This holistic approach to data analysis empowered the non-profit to make data-driven decisions and foster stronger connections with their audience.
This project uses Tableau to create an interactive dashboard with Airline Schedule data. It offers a map displaying outgoing flights from a selected city, along with carrier information, destinations, and seat capacity. Users can also select destinations for flight duration and aircraft amenities, or choose an airline to access details about all destinations, travel information, and in-flight services.
I am currently working in Siemens Energy to develop a domain-specific Cognitive Assistant to streamline information retrieval for employees, enabling faster access to relevant data without the need to manually sift through extensive documentation. Prior to this, I have completed three internships as a Data Scientist/Analyst, during which I honed my skills and provided valuable support to companies in their data-driven decision-making endeavors.
Before my internships, I worked as an SAP ERP consultant, acquiring insights into the inner workings of corporate processes and understanding how different organizations collaborate to achieve their business objectives.
Linkedin Download My ResumeSecured 3rd position among 20 teams in the "Reinvent the Wheel 2.0" hackathon organized by Torqata and American Tire Distributors. This competition was about leveraging Data Science to develop sustainable solutions that reduce the environmental impact of the tire and automotive aftermarket industry.
This recognition acknowledges my consistent dedication and exemplary performance. The award was granted in recognition of my remarkable track record in consistently achieving Service Level Agreements (SLAs) and consistently delivering high-quality work on time with an exceptionally low rate of rework.
This was a recognition to highlights my commitment to technical excellence and best programming practices in project implementation. It underscores my dedication to employing the most efficient programming approaches to deliver high-quality solutions. It serves as a testament to my focus on continually expanding my technical expertise.