Data Science Intern

• Jan 2023 - Present

Data Scientist

• Mar 2020 - Aug 2022

Created an end-to-end model retraining pipeline with Luigi to retrain, evaluate and serve updated models in an automatic and periodic cadence. Worked with Cross-Functional teams on A/B testing model updates, researching zero-day Phishing attacks, monitor model health, and developing internal tools for data curation, visualization, and analysis Created an end-to-end model retraining pipeline with Luigi to retrain, evaluate and serve updated models in an automatic and periodic cadence. Worked with Cross-Functional teams on A/B testing model updates, researching zero-day Phishing attacks, monitor model health, and developing internal tools for data curation, visualization, and analysis

Data Scientist

• Oct 2019 - Mar 2020

Reduced a significant amount of effort put in by underwriters to assess the risk of insuring people by clustering insurance documents using Universal Sentence Encoder and K-Means. Reduced a significant amount of effort put in by underwriters to assess the risk of insuring people by clustering insurance documents using Universal Sentence Encoder and K-Means.

ADAS Intern

• Feb 2019 - Jul 2019

Worked on a POC implementation in Level-3 Automated Driving Systems – developing an ensemble Deep Learning architecture to detect Construction Sites well in advance. Worked on a POC implementation in Level-3 Automated Driving Systems – developing an ensemble Deep Learning architecture to detect Construction Sites well in advance.

Machine Learning Intern

• May 2018 - Jun 2018

Provided quicker support for end-user queries in one of India’s leading banks using a Q&A chatbot based on TF-IDF and Cosine Similarity. Worked towards supplanting the retrieval-based chatbot with a pre-trained generative chatbot based on Seq2Seq model, sourcing data from Cornell Movie Dialog Data corpus, containing 200k+ conversational exchanges. Provided quicker support for end-user queries in one of India’s leading banks using a Q&A chatbot based on TF-IDF and Cosine Similarity. Worked towards supplanting the retrieval-based chatbot with a pre-trained generative chatbot based on Seq2Seq model, sourcing data from Cornell Movie Dialog Data corpus, containing 200k+ conversational exchanges.

Python Developer

• Jan 2018 - Apr 2018