Hi, I'm a computer science and engineering graduate who's been working as a Data scientist since September 2019. Prior to that, I was a Machine Learning Intern at Suprath Technologies from November 2018 - March 2019.
While normally one wouldn't consider internships as "proper" work experience, this particular internship was rather fleshed out, my code made it to production. My thesis summary is also uploaded, in the event that you would like to see the full text, please do email me at deeptanshupaul@gmail.com. At Suprath Technologies, I learned a lot. I went from learning about ML theoretically to actually understanding the nuances of making production-ready Machine Learning systems. Not only that, striking the balance between academic rigor and industry requirements in performance was another major challenge that I overcame here. The key technologies used in this project are Python (Keras with Tensorflow & other basic libraries like nltk, pandas,etc), Facebook AI Research's Fasttext vectors & Stanford's GloVe vectors.
My bachelor's thesis was titled as "A Percept-Affect Model of Empathy in Conversational Artificial Intelligence" and it aimed to bring in a degree of emotion recognition (percept) and response (affect) to a conversational agent. Using neural networks to detect emotions, as well as context to detect sarcasm, the system modulated responses across 5 levels of formality to give relevant responses to percepts. This was a solo project done under the supervision of Aman Rusia (the then CTO of Suprath Technologies) and Professor Sujatha AK of Christ University, Bengaluru.
As a Data Scientist at Perpule (Delvit Solutions), my primary directives were to standardize the various practices with regards to analysis of data (Data Governance), the transformation of data for quick analysis, the actual analysis and visualization of data. I worked on setting up pipelines, designing databases and then moved onto the generation of insights from said data. Since Perpule is an e-commerce company, the key application of data science comes in the way of predicting sales, supply requirements (inventory planning) and demand, and extracting patterns in shopping data. Market Basket Analysis was one such project that was deployed. In addition to Data Science, I was closely involved with the design and development of a new product, both in from a technical perspective as well as a managerial perspective. I took a lot away from this experience, understanding the intricacies of major product design decisions was one of the big learning experiences that I experienced.
Please find my Resume as well as a summary of my thesis in this repository. I've done several projects in college, and have always approached Machine Learning & Deep Learning from a mathematical perspective. An example of my work in college when it comes to explaining topics in ML.
"What I cannot create, I do not understand" - Richard Feynman
Apart from the projects listed in my Resume, almost every topic I've studied in the field of AI has a practical implementation as well.
Thank you for your interest.