Aditya Rawat

Work History

Data Scientist 2

Sixt

Munich

06.2022 - Current

Built a price recommendation system for Car Rentals established on Catboost Model and RBFOpt Optimizer. Catboost predicts the demand based on prices (< 2 MAE Score)and Optimizer fit the demand to an ideal utilization curve by finding the right price.
Deployed the Price recommendation system on AWS.
Built automated analysis tool for Online Marketing team to analyze Ad Campaigns running on Google saving 2-3 hours of labour per Day.
Gained strong leadership skills by managing projects from start to finish.
Actively participate and contribute to the internal data science project initiatives

Data Scientist

trivago

Düsseldorf

12.2020 - 05.2022

Working within B2B side of trivago with Account Managers providing data-driven Insights helping Advertisers run their campaigns successfully on trivago's Marketplace.
Developed and Deployed Hotel and Bid Recommendation Model for finding well-performing Hotels for Advertisers along with Bids for Sponsored listings Auction.
Designed and deployed KPI Monitoring Dashboards using looker.
Set up reporting workflows to automate reporting processes saving 24 hours of labor per month.
Training and support of new team members.

Data Scientist

Hughes Systique Corporation

Gurugram

07.2018 - 11.2020

Footfall Analytics: Deployed a Footfall prediction model for TJ Maxx, American store chain, on GCP using Cloud Function and Improved prediction rate by 68% using FbProphet. Learned how Social Marketing Campaigns and Weather Type affect footfall in different regions.
Image Installation Auditor tool: Image Classifier based on VGG16 model, deployed on GCP for Image Auditing. Increased Auditing efficiency by 95% by automating classification of different installed hardware equipments.
Inventory Management System: Based on Monocular Depth Estimation determined with 98% accuracy number of bottles inside commercial refrigerator (A POC for Coca-Cola India).
Measuring Retail Store Traffic System: Live tracking visitors inside store with 81% Multiple Object Tracking Accuracy (MOTA) using Person Re-Identification Algorithm (feature map extraction and Pedestrian Attribute Recognition) with Siamese Neural Network.
Call Center Transcripts Generation: Used Speech Signal Processing and Speaker Diarization to distinguish between Customer executive and caller, Speech Recognition to convert speech into text. Achieved 25% WER(WordErrorRate).
Text Summarization and Sentiment Analysis: Developed a tool that summarizes conversation between Customer Care agent and Caller for Hughes Network Systems and then rates conversation using sentiment analysis. Based on extractive text summarization and used ELMo Model for sentiment analysis.
Unsupervised Text Analytics: Developed approach to find different intents, hidden inside a sentence, question, or comment in unsupervised manner using PCA and clustering. Raw sentences are encoded using Sentence Bert (SBert)and projected onto top 100 Eigenvectors using PCA capturing 95% variance. Performed cluster analysis to find hidden intent clusters. An optimal number of clusters was determined using WCSS and elbow method.

Freelance Data Scientist

PeoplePerHour

Delhi

01.2019 - 05.2019

Profile can be viewed at pph.me/addie11
Provided expertise in data mining and data cleansing.
Interpreted data in order to draw conclusions for insightful action and strategy using interactive visualizations for Alemeno.com, start-up that uses behavioral science to help kids inculcate healthy food eating habits.

Summary

Overview

Work History

Data Scientist 2

Data Scientist

Data Scientist

Freelance Data Scientist

Education

Bachelor of Technology - Information Technology

Skills

Certification

Work Availability

Timeline

Similar Profiles

Saleh AlghamdiSaleh Alghamdi

Jared Estrada-LucianoJared Estrada-Luciano

Krystal BrittonKrystal Britton

Devkumar BeharryDevkumar Beharry

Olha HudzenkoOlha Hudzenko