Feb 2023 – Mar 2023 · 1 mo

• Voluntary Research Assistant , SUNY University at Buffalo 02/23 – 03/23
• ● Assisted Prof. Lesli Li in her project called Toxic Comment Detection using the GATE model.

Aug 2022 – Jan 2023 · 5 mos

• Programming and Database Fundamentals for Data Scientists

Aug 2021 – Feb 2023 · 1 yr 6 mos

May 2022 – Aug 2022 · 3 mos

• ● Built a Spring Boot service Authorization and Authentication Module for Consumer org. using ECS Fargate AWS Infrastructure.
• ● Added data transfer security by end-to-end TLS using Network Load Balancer.
• ● Set up AWS PrivateLink which connects VPC of AWS services without exposing traffic to the public internet.
• ● Built CI-CD data pipeline which includes developing, packaging, testing, validating, and automating 50% of workflow deployment.

Apr 2021 – Jul 2021 · 3 mos

• ● Developed End to end conversational AI agent for money transfer across countries using BERT Model.
• ● Improved intent detection accuracy by 9.2%; had a positive impact on 100,000 customers and did data analytics of chats.
• ● Automated testing of the bot using Scrapy, BeautifulSoup, and Python optimized 72 hours of manual effort.

Nov 2018 – Mar 2021 · 2 yrs 4 mos · Pune Area, India

• Talentpool Chatbot
• ● Developed a customized recruitment chatbot with 94.6% accuracy of intent classification using BERT.
• ● Centrality Analysis of Graph DB (Grakn) to capture 46% of user intents built on Tensorflow framework.
• ● Automated deployment using MLOps and testing using CI/CD pipelines leveraging Github Actions and AWS cloud
• Personal Protective Equipment Detection
• ● Implemented a Safety object detection system for construction workers using Tensorflow framework
• ● Model training using CNN based Inception V3 architecture and calculation of safety index using Tensorflow framework. Construction Management Application, Autodesk
• ● Designed custom microservice architecture for project management using the Django Rest framework.
• ● Developed independent and decoupled services allowing a more agile approach to software development.
• ● Leveraged asynchronous framework based on Kafka queuing service to optimize wait time to 2ms.

Aug 2017 – Oct 2018 · 1 yr 2 mos · Bangalore

• Social Sensing of a Product – an Advertising platform.
• ● Created a SpringBoot Application, with data streaming from Twitter to Kafka and used HDFS storage using Spark.
• ● Analysis of the publicly available social media data to understand the outlook of smartphone brands using a Prediction model using
• tensorflow and Spacy libraries using NLP techniques using Pyspark, Kafka, Zookeeper.
• ● Training pipeline involves Text preprocessing, TF-IDF Vectorization, Modelling, and Testing.
• Text Summarization
• ● Created a SpringBoot Application, with data streaming from New websites to Kafka and streamed in HDFS system
• ● Implemented an extractive algorithm that aids in summarizing textual information such as news articles, chats, etc.
• ● Based on key factors like text length, thematic weight, nouns, sentence position, numeric data, and term weight.
• ● Accuracy was measured by gold standards BLEU and DeFUSE.
• Cognitive Search Engine
• ● Scalable semantic search engine to handle natural language queries from both structured and unstructured sources.
• ● Trained AI Watson to understand the terminologies of the oil industry using Watson Knowledge Studio.