Sourish Dey

Lead Data Scientist @Centric Software

💸 Marketing✔️ ITML EngineerMarketing Analytics Manager🧑‍🔬 Engineer🧨 ConsultingData EngineerComputer engineerBusiness AnalystData ScientistCustomer ResearcherData Analyst

With more than 11 years of experience, Sourish Dey is currently serving as the Lead Data Scientist at Centric Software, a global leader in providing innovative AI-ML driven solutions for fashion, retail, footwear, outdoor, luxury, and consumer goods companies. With a strong background in machine learning, data mining, and predictive modeling, Sourish is instrumental in leveraging vast amounts of structured and unstructured data to drive strategic business decisions and build AI capabilities. Prior to joining Centric Software, he held various data-centric roles at several companies in different industry sectors(Financial Services, Retail, Manufacturing, eCommerce), like Siemens, GE, Citi, TCS,Delivery Hero, dunnHumby etc. Sourish holds a Master's degree in Business Analytics and Econometrics and Bachelors in Electrical Engineering and continues to stay abreast of the latest industry trends to ensure his proficiency in the field. He is known for his ability to communicate complex AIML solutions in a clear and easy-to-understand manner.

Feedback from Sourish’s ex-colleagues

Let me see

Keep in touch with meI'm using Intch to connect with new people. Use this link to open chat with me via Intch app

Work Background

Centric SoftwareSr.Data Scientist

Nov. 2023Berlin, Germany

NEW YORKERSr.Data Scientist

Mar. 2023 - Oct. 2023Berlin, Germany1) Developing, and deploying Assortment Size Optimization in global item(12 K SKU)-store(1150) level and Monte-Carlo simulation of sales to realizing EUR 0.2M/year of additional sales and cost-saving EUR 0.3M/year. 2)Conceptualizing and professionalization prediction model for new items leveraging deep learning and transformer methodology using computer vision and NLP to reduce 9% of inventory cost.

Delivery HeroData Science Specialist

Oct. 2021 - Oct. 2022Berlin, GermanyAutomatic Voucher Optimization, Marketing Uplift Model, Time Series Forecasting Development and productionalization(AWS,Docker) of M/c Learning based(Causal Uplift Model) Automated Voucher Optimization product(optimizes voucher distribution at customer level) in to maximise incremental customer acquisition and engagement in budget constraint for FoodPanda, APAC.

dunnhumbyLead Research Data Scientist(Machine Learning and Recommendation)

May. 2019 - Sep. 2021Berlin, Berlin, GermanyData Science | M/c. Learning | Business Analysis | Big Data | Deep Learning Developing predictive models and personalized recommendation algorithm for real time bidding and ad viewing, leveraging supervised and unsupervised ML-DL techniques for world's first customer data science platform. Working closely with business owners and S/W engineering(production and DevOps) teams to solve challenging business problems in eCommerce,retail and media(advertising technology) industry.Data Science research and solutions to revolutionize the process and user experience to create dunnhumby media AI solutions.

CitiSr.Data Scientist

Mar. 2017 - Apr. 2019New York City Metropolitan AreaData Science | Machine Learning | P&L Business Analysis | Big Data | Deep Learning Enhancing Digital Targeting (over 300 MM prospects) using scalable next-gen machine learning and deep learning solutions for GCB(Global Consumer Bank) at GDM- NA Cards- Digital Acquisition. 1) Developed and Implemented Product Specific Response Model for direct mail acquisition channel(North America Credit Card business)using XGBoost technique within 2 months. Further research in the expansion of model feature space using CNN(Convolutional Neural Network) technique and model performance enhancement using deep learning technique. 2) Building m/c learning models(propensity to convert/click, wallet segment models etc.) for Citi's digital acquisition decision flow engine for customized digital targeting and offer display in different digital channels(display media, paid search,social media) to maximize post-conversion CMV. 3) Developed ML based Balance Transfer amount model for Citi's proprietary Branded Cards products using GBM technique. 4) Currently working on building predictive response models using M/c learning techniques(GBM, Random Forest)in the area of digital email campaigns for cross selling(add a card) Citi's proprietary products. 5) Building M/c learning(SVM,RF,XGBoost) and deep learning(ANN) based prototype solution for proactive customer retention for Citi's phone banking channel. Tools Proficiency: A) Scripting language - R/Python B) Big Data- Hive/ Tectia/Spark(PySpark,RSpark) C) Cloud Computing - AWS, Floyhub D) Data Analysis Tool- SAS,SQL Techniques Used: A) Statistical Techniques: Logistic Regression, Regression,Cluster Analysis B) Machine Learning Techniques: GBM, RF, XGBoost C) Deep Learning Techniques: ANN,Convolutional Neural Network(CNN)

Tata Consultancy ServicesData Scientist

Dec. 2015 - Mar. 2017Mumbai Area, IndiaData Science| Machine Learning | Statistical Modelling| Risk Analytics| Consulting 1) Leading (with individual contribution) a 5 member team of Data Science and Big Data in advanced analytics consulting ( Requirement Generation, Requirement Analysis etc.), developing end to end data science solutions across domains and delivering results to the stakeholders. 2) Responsible for delivering value in different business functions (Risk,CRM, After Sales, Marketing etc.) through data science projects and POCs. Business Analysis and Developing predictive models using statistical methods(Statistical Distributions, Regression Methods, Decision Trees etc.) and new age Machine Learning Techniques(Deep Learning, XGBoost, Random Forrest, Neural Networks etc.) 3) Mentoring Jr.Data Scientist and statisticians in solution architecture and Knowledge Sharing. Tools Proficiency: A) Data Science - R/Python/SPSS B) Big Data- SPARK/ PostgreSQL C) Visualization - Tableau,QlikView

SyntelLead- Analytics and Data Science

Dec. 2014 - Nov. 2015Mumbai Area, IndiaBusiness Analytics| Machine Learning | Risk Management| Project Management Current Assignments: 1) Currently engaged in Analytics development and deployment project for US P&C insurance giant Allstate (Integrated Claim Analytics), providing business insight in the area of Claim loss, Claim Severity, Agency, Perils, Coverages, Products to identify multivariate high risk pockets, building claim propensity predictive models, claim risk profiling and subsequently building Risk scorecard and behavioural scorecard helping business in underwriting and policy renewal. 2) Exploratory and advanced analytics in Fraud detection and pricing and revenue analysis(CLTV, price elasticity of demand etc.) for Allstate. Data Mining and Machine Learning: A) Classification : Logistic Regression, SVM, Decision Tree(CART, RPART, CHAID), Naive Bayes B) Prediction and Forecasting: Linear Regression, Time Series Analysis(ARIMA, Holt Winter,GARCH) C) Clustering and Variable Reduction Techniques: K-Means, Hierarchical, Variable Clustering, PCA Tools Proficiency: A) Data Extraction and Preparation: Hive-SQL, B) Machine Learning and Statistical Modeling - R/R Studio/Rattle, MATLAB/OCTAVE,SAS C) Visualization and Productionalization - Tableau