Developed chatbots using Python, RASA, BERT, and other deep learning models like Word2vec and Doc2vec. Also, made the speech-to-text noise robust ensemble model using DeepSpeech and RNN- Noise. ● Worked extensively with python, deep learning models and used SQL to analyze the conversa- tions to make the chatbot more conversational. ● Used sequence to sequence models for better transcription of audios even in the noisy environ- ments.

Got full scholarship to attend Pi School of AI in Rome, Italy where I along with my partner made a ma- chine learning tool that does AWS resource optimisation. The cost of AWS infrastructure was reduced by 20% by implementing the predictions of the tool. ● Converted the business problem of ENEL to a solution which was very valuable for them after communicating with the stakeholders. ● Used data analysis and data engineering skills to deal with 450 GBs of messy data and get valua- ble insights from it. ● Came up with machine learning models after feature engineering which were 92-94% accurate in predicting which instance should be downsized, stopped or terminated based on the metrics of instance and the nature of project executing on them.

Worked on two research projects namely "Classifying Alcoholics and Control Patients using Deep Learning and Peak Visualization Method" and "Textile Designs Generation using GANs". Both were ac- cepted for presentation and publication at peer-reviewed conferences i.e. WSPML 2019 and 33rd IEEE

Completed 65+ freelancing projects of data analysis, data visualization and modelling with Python and machine and deep learning models with an average rating of 4.9/5.0. I am a level 2 seller at Fiverr. ● Communicated with clients and turned their business problems into development problems and implemented solutions that have value for them. ● Gained hands-on experience on numpy, scipy, scikit-learn, pandas, matplotlib, tensorflow, keras, pydub, openCV, and other libraries that assist in making the projects.