I am a Full Stack Data Scientist, working to help businesses grow exponentially using the power of data.
I have worked on an array of end-to-end Data Science projects, interacting with clients, setting up the specifications and delivering and deploying the final solution within their premises. I aim to bring the same quality of work to all my clients, and leverage my understanding of Python, and it's utility in various ML and Deep Learning Pipelines, with your data, to deliver actionable insights.
Having worked on a range of projects across domains - Customer Support, Sales & Marketing, I understand the business' requirements from the data-team, and can help establish up data driven processes, and scale existing ones.
Some of my career highlights are -
1. Created a NLP based classification solution to classify incoming support requests for a US based healthcare giant, leading to a 45% reduction in resource utilization and faster
response time.
Technologies - SQL, Word2Vec, Multi-Layered Perceptron, Random-Forest Classifier, TFIDF
2. Scraped data from various social networks, to generate leads for marketing and increase user registration by a factor of over 50%. Additionally, automated mundane tasks, such as sending out personalized messages, to free up employee time and increase productivity.
Technologies - Selenium, BeautifulSoup4, python-requests
3. Developed a python based tool to perform text analytics on live social media(Twitter, News, Facebook, Blog) comments to understand user's view of an organization and reduce negative sentiment among them.
Technologies - Django, BS4, tweeps, K-Means Clustering, VADER
4. Created a rule based engine to identify various entities in customer feedback comments and determine the sentiment linked with each entity to better understand the customer pain points and prescribe areas of improvement to the business.
Technologies - SpaCy, StanfordNLP, VADER, TextBlob, K-Means Clustering
Work Terms
* Hours of operation - Generally 700 - 2300 IST, but can be moved around a bit to suit the client's requirements
* Prefer being contacted over mail/text
* Not here to low-ball and get into bidding wars. I will give you quality results and in an efficient manner