Nov 2020 – Mar 2021 · 4 mos · Bangalore Urban, Karnataka, India

• ● Successfully productionized a new agnostic OCR system for parsing invoices of different procurement companies to reduce the time spent on basic analysis done by the operations team at the end of each quarter.
• ● Deployed an engine to convert news reports read on channels to full-fledged news articles using T5.

Jan 2020 – Nov 2020 · 10 mos · Bengaluru, Karnataka, India

• ● Worked on creating content filter with the application in healthcare insurance industry for hiding PII and sensitive information in diagnostic reports with a macro F1 score of 0.96.
• ● Successfully completed the delivery of PII filtering project deployed in an end-to-end Azure ML Pipeline, supported with MLOps features like feature store, data drift, Explainability.

Jun 2019 – Jan 2020 · 7 mos · Bangalore

• ● Delivered Intent Classification Engines using state of the art NLP techniques like BERT with fine-tuned deep learning model.
• ● Successfully deployed models with Transfer Learning techniques to increase accuracy of basic NN models.
• ● Production grade, Machine Translation using Seq2Seq (encoder-decoder) NN models.
• ● Handed over full-fledged conversational assistants for banking and appointment domain in Indic Languages using RASA NLU and Core.

Jul 2018 – May 2019 · 10 mos · New Delhi Area, India

• ● Deployed, Location Based Assistant with Intent based classifications and Named Entity Recognition using SpaCy.
• ● Responsible for completion of, Autosuggest Engine for category-based Location Searches: Trie Based Information Retrieval.
• ● Spell Checker Using SCRNN, currently helping search algorithms.
• ● Research experience on Unsupervised Analytics of API hits and running mean-based anomaly detection.
• ● Quick chatbot prototypes in Amazon Alexa with Python SDK.