Jan 2022 – Nov 2024 · 2 yrs 10 mos

• Automated personalized marketing content generator| Healthcare | @Bangalore
• As an engineering manager, I led a team that worked closely with client counterparts to develop and deploy a generative AI system to create customized marketing materials. This technology streamlined the production of unique banners, emails, and digital service platforms for various U.S. brands, enhancing customer engagement through personalization.
• Directed a multidisciplinary team comprising backend engineers, data scientists, and frontend developers to pioneer a novel and highly dependable genAI-based product, marking a first in the company's history. This product sets a new standard for reliability, availability, and precision in AI-driven solutions. Orchestrated the deployment of a production-grade API for conversational chat, accompanying it with comprehensive API documentation to ensure seamless integration and utilization.
• Designed and implemented a strategic pricing strategy for the API, ensuring value-driven monetization while meeting market demands which commands a savings of more than 8M dollars by 2025.
• Led the development using various tech stacks such as reactjs, springboot, azure openai hosted in AWS using services such as Kubernetes, lambda, cloudfront and istio.
• Chatbot API Development for Consumer Plant Queries| Non Profit | @Bangalore
• Collaborated closely with the RHS team to gather, comprehend, and refine project documents, ensuring alignment with desired outcomes.
• Pioneered the creation of an endpoint API tailored to cater to plant-related inquiries for the chatbot, leveraging client-provided data and references.
• Orchestrated the deployment of a production-grade API for conversational chat, accompanying it with comprehensive API documentation to ensure seamless integration and utilization.
• Designed and implemented a strategic pricing strategy for the API, ensuring value-driven monetization while meeting market demands.

Apr 2019 – Dec 2021 · 2 yrs 8 mos

• Demand Forecast | Retail | @Dubai:
• Facilitated development of an automatic ordering system (AO) by coordinating with multiple stakeholders across different geographic locations contributing towards the revenue by 2M AED yearly (reduction loss of sales).
• Designed machine learning solution leading to the development of highly scalable demand forecast model(s) to predict future sales of over 2000 SKUs.
• Designed experimentation strategy to measure supplier service level. Observed a 10% rise in supplier service level during MVP phase.
• Additional: Developed an early warning monitoring and notification system to predict the item being not on the shelf using transactional data. Designed experiments that led to a 26% decrease in Not-On-Shelf (NOS) rate. Blog
• Reviews dynamic visualisation (NLP) |Travel and Hospitality | @London:
• Created a dynamic visualizer to concisely investigate the performance of apps using unstructured reviews text.
• Deployed NLP pipelines to aggregate data, build models and inject the output to dynamically update dashboards hosted using streamlit.
• Data Science Architecture: Architecting and building Data Science/Machine Learning infrastructures, solutions, pipelines, and models by implementing engineering patterns, and automated workflows.
• Customer Next Best Offer: Developed and deployed a near real-time data-driven model to provide the next best product offer to customers to boost x-sell/upsell activities.
• Customer risk: Created a batch prediction suite of models to predict customers at risk of lapsing and surrendering.
• Extended behavioral segmentation model website to identify intent and status of path-to-purchase funnel enabling company to cut down million visitors to thousands of potential buyers.

Oct 2016 – Mar 2019 · 2 yrs 5 mos

• Customer Next Best Offer | Insurance domain | @Singapore:
• Collaboratedwithbusinessteamstobuildthenextbestproductandriderofferingtoexistingcustomer base to boost x-sell and upsell activities.
• Builtahybridrecommendersystemusingcustomerdemographicsandrelevantproductfeatures obtaining a lift of 15% AUC compared against the popularity-based model.
• Hyper-personalizationledtoapotentialrevenueuptickof~$1MSgperquarter.
• Reviews dynamic visualisation (NLP) |Travel and Hospitality | @London:
• Created a dynamic visualizer to concisely investigate the performance of apps using unstructured reviews text.
• BuilthighimpactNLP-basedproductfeaturingTopicexplorer,wordexplorer,relatedconcepts, emotions explorer, sentiment polarity using unstructured data scraped from google play store.
• Deployed NLP pipelines to aggregate data, build models and inject the output to dynamically update dashboards hosted using streamlit.

Dec 2014 – Sep 2016 · 1 yr 9 mos

• Analysed NIGO (not in good order) volumes and proposed more than 31 recommendations to the business to restrict these types of unwanted volumes in order to improve customer satisfaction score and associate work experience.
• Forecasted arrival volumes (both short term and long term) weekly and monthly basis to be used for workforce optimization.

Nov 2013 – Nov 2014 · 1 yr

• Created an indicator called Health score index to capture the health of a process in XTRAC – a tool which captures end to end workflow of requests.
• Automated report generation process for Short term and long term analyses using VBA, Python and Tableau.

Oct 2012 – Oct 2013 · 1 yr

• Led statistical coding team for the same with a focus on creating automated scripts to perform data cleaning, manipulation & modeling.
• Researched on new features and designed GUIs and flow for the same.
• End to end engagement with development team for successful deployment of the new feature.

Jun 2012 – Oct 2012 · 4 mos

• Part of the core team in development of in-house analytics products.
• Successfully deployed multiple features in a product using Python & R in an on-prem cloud environment.

Jun 2011 – Sep 2011 · 3 mos · CORMSIS Labs, Southampton, United Kingdom