Utkarsh Harit
@utkarshharit
Data Science Consultant at ACCENTURE (solutions.ai)
Jaipur, India
Data Scientist with strong technical background and experience in working with predictive modeling, data processing, and data mining algorithms to solve challenging business problems. Highly skilled at Machine Learning, Deep Learning and Data Science and passionate about data. Interested in working on challenging and complex problems through creative thinking and research.
Experience
Data Science Consultant— Data Scientist
ACCENTURE (solutions.ai)
Build insights, high impact data models and act as a subject matter expert. Fine-tune of base GPT-3 model, focusing on enhancing Gen AI accuracy and ensuring consistent outputs on the Azure to improve NLP outcomes. Developed a solution for a NLP based research problem for clients to understand and improve employee’s health and experience. Acquired multiple clients by building solutions for the retail industry to extend audience reach by creating KPIs using recommendation engine, RFM model and statistics on Cloud which were dockerized and deployed on the client's GCP and Azure system. Saved 90% cost by Recommending sales motions for a client with a vast product line. Tech Used - Python, Deep Learning, Data Science, NLP, Azure, GCP, ML, Data Proc, Databricks, Azure SQL, BigQuery
Assistant Manager— Data Scientist
Vaibhav Global Limited
Development, Deployment, Maintenance and Evaluating Benefits of AI based solutions for ShopLC (USA) and TJC (UK) to solve business problems with Predictive Modeling, Data Analytics, Statistical Modeling, Automation and Managing Data Science Lifecycle on Large Datasets. Accomplished 85% accuracy for NLP task of Customer feedback and Sentiment Analysis by using Bi-directional LSTM. Accomplished 90% accuracy for Image Recommendation system for searching similar products from image by using Autoencoders and KNN. Optimization for product scheduling on TV for 24 hrs by using the LightFM recommendation system. Product Comparison tool for comparing product with the competition in the market to acquire new users, retain existing customers by collecting data from 50+ websites using Scrapy framework. Replacement of manual data collection with Competitor Analysis tool for Trend analysis and best product analysis by data gathering from 15+ different Data Sources using Requests. Detecting the effect of Channel Anchor’s expression and speech on sales from live video by Predictive Analysis with Emotion Analysis Tool. Developing a GAN based Jewellery Design Generation model. Tech Used - Python, Deep Learning, Machine Learning, Data Science, NLP, Computer Vision, Web Scraping, Bitbucket
Intern— Data Science Intern
Forsk Technology
Provide Business Solutions to the Forsk Client, Product Development by creative planning and attention to detail, Identifying risks, formulate strategies and Recommending Solutions. Classification of product by hypothesis testing of using Image Features and Product Features as collaborative input into ANN model. Tech Used - Python, Deep Learning, Image Processing, Web Scraping
Research Intern— Scholar
Florida International University
Developed a IOT based project, to demonstrate the difference in calories burnt as the temperature changes, under the technical and resourceful guidance of Dr. Shekhar Bhansali. Tech Used - Python, Internet Of Things
Education
Anand International College of Engineering
Bachelor of Technology
Computer Science Engineering