KAUSHIK SHAKKARI
Hello, I'm Kaushik
a senior data scientist, author, and mentor based in Seattle, currently working at skills-matching intelligence based Startup, AstrumU. Previously led SQUARE, an NLP search product at Cognistx.
Experience
I worked on building Machine Learning and Deep Learning pipelines for companies, including Cognistx, Symantec, Cognii, and Cisco. I also worked on Web Browsing Behavior Analysis and Health Information in research labs in the United States and India.
Senior Data Scientist
Enhancing skill extraction process by enabling document parsing capabilities within the AI Translation Engine, enabling image and text parsing of semi-structured documents like transcripts and resumes (curriculum vitae), thus augmenting segmentation, detection, and recognition functionalities.
Projects
I have worked on variety of different interesting data driven projects.
Self Driving Car
Modeled a self driving or autonomous AI agent trained on GTA 5 game. Object Detection (cars, humans, traffic signal etc) and Collision Detection are successfully implemented. Now, working on shifting the model to a hardware prototype which has Raspberry Pi 3 B+ OS using transfer learning. The hardware prototype is equipped with a 5 MP 1080p camera.
Achievements
I am happy to share that some of my works are recognized by companies and universities.
Ongoing Patent
I have submitted the IDF (Invention Disclosure Report) to the Symantec Patent Committee. Symantec patent committee has reviewed my idea and work and decided to patent it. Currently working with external law firms to file the patent.
Patent Domain: Security for sensitive data
Technical Domains: Natural Language Processing, Deep Learning and Transfer Learning.
Data Science Case Studies
I am always interested in solving data driven problems. Most of the case studies are given by professors from different departments like business, health informatics and environment. I have investigated on
problems and presented my solutions to them.
Analyzing product and developing pricing and product strategy
The goal of this project is to understand how online data can improve pricing and product strategy. Interactive charts and plots are developed to extract insights from data. The dataset consists of sales of tablets sold by different companies like Samsung, Apple, Kindle and others. Each product has its sales rank and other attributes like price, processor speed, discount percentage and average rating given by customers. In this analysis, I first understood the factors that impact the demand of tablets in the market. Another goal in this project is to increase the sales of Samsung tablets by studying the relationship of the variable sales rank.
Co-Curriculars
Campus Ambassador
I have got selected from my university and worked as a campus ambassador for one of the India's biggest computer science portal and got certified from the organization.
Roles and Responsibilities:
1. Digital Marketing
2. Conducting workshops, Geek Classes and seminars in the university.
3. Motivate people from different domains to practice computer science