Client facing data scientist who is passionate about his work and enjoys solving real life problems by leveraging data & unfolding its story.
I have gathered tremendous experience over various industries, due to client exposure and development over the last 2+ years, working for Webhelp’s blue chip clients such as major airlines, food delivery platforms, banks etc. I have so far mainly focused on the NLP tech, to name a few, BERT, GPT3, token classification, sentence classification, sentiment analysis, emotion detection, topic modelling, translation, audio-to-text. This involved also creating pipelines for performing transcriptions as well as applying machine learning algorithms and producing valuable insight. Everyday basis is based on utilizing Azure Databricks for accelerated computing alongside big data frameworks such as pyspark, deep learning frameworks like pytorch and transformers, and more according to business needs. Example projects include producing an AI toolkit, which enables to accelerate call assessments, assess advisors based on performance, attitude and communication skills ; which once implemented, saved thousands of ManHours I like to build relationships and communicate in non technical terms, simplifying as needed to make sure my audiences fully understand my work and the resulting outputs Before my permanent role at Webhelp, I've created a text classification engine, as part of my industrial placement with them, alongside my Master's degree. This tool was a stepping stone for them as before that they concentrated more on pure analytics instead of NLP ; this led to quite a lot of additional projects which I took on and led when I worked for them full time after my studies. I would also like to mention my 4-year's bachelor thesis which included Deep Learning, using LSTMs to track audio music, translating it to midi, on-line, as the music played.