Shishir Kumar

Background

About Shishir Kumar

My journey from a Full Stack Engineer to a Machine Learning Engineer. I started my career as a full-stack engineer in 2013. From the initial days of my career, I have been interested in visualization and building cool widgets. After building several visualization and insightful dashboards, I became interested in telling more stories through the data. I took up the Data Analyst and Data Engineering role to explore different kinds of domain data, build data pipelines in the Cloud and create a variety of dashboards. After working in the Analysis of domains like Social Analytics, Ad Analytics, Legal Analytics, and Smart Energy Analytics, I pursued my interest in data to the next level by building Machine Learning Models. I have completed my Post Graduate Certification in Machine Learning and Natural Language Processing from Upgrad and Bangalore with an excellent academic record in Statistics and Machine Learning (4 CGPA) and NLP (3.70 CGPA). I am pursuing my Master's in Machine Learning and Artificial Intelligence from Liverpool John Moores University. I worked at Fokal AI for 4.5 Years as a Machine Learning Engineer and have built different interpretable ML models to help our clients with actionable business decisions Fokal AI. I am greatly interested in the NLP domain. I have worked on Text Generation, Text Classification, Sentiment Analysis, Emotion Classification, Intent Classification, Effort Classification, Information Extraction from Text Reviews, Text Summarization and Question Answering systems. I have strong experience deploying Machine Learning Models in Data Pipelines or Serverless architecture. I have experience with AWS, GCP and Azure. In January 2021, I took up a new role as a Machine Learning Engineer at Boast.AI and migrated to Vancouver, Canada. I am currently working on LLM / NLG - Text Generation, Text Summarization, Text Classification, Clustering and building ML pipelines. I have been in absolute love with the Large Language Models since 2021. Using the latest technologies, I have had a chance to build many LLM applications for Report Generation, Conversational AI, Chat Applications, Document AI, Long-form and Short Form of Text, Hybrid Search, Question and Answering Systems, Summarizing reports, etc. Working with Large Language Models, I have built advanced RAG Applications to reduce LLM Hallucinations, fine-tuning generic LLM and Open Source LLM using techniques like In-context Learning, Few-shot Learning, and Transfer Learning. I am on LinkedIn to grow more, meet people and follow LLM / NLP / ML / AI topics.