Kiran Hegde

Humbled to be serving as Technical advisor to the CEO, a role I reckon I will learn more than I can advise. Work includes overseeing (1) Partnering/overseeing enterprise level AI Initiatives at Chewy (2) Looking at internal and external AI trends and helping craft a cross-dept strategy, (3) Creating a vision/strategy for new initiatives (where needed) and creating alignment between different groups. Deliverables include 1. Lead Agentic AI Platform Product & Tech vision & led home grown build enabling orchestration for Agent to Agent interaction, observability, security, and data governance for AI agents across key domains, boosting Agent build productivity by ~$0.xM+ per domain 2. Lead Enterprise MLOPs Platform to accelerate alignment, standardization and adoption of tools. The MLOPs practice is spread over 3 key areas (1) Feature Engineering – which turning raw data into useful signals, consists of Data Retrival and Prep (2) Model Building – involves looking at past data to predict future, consists of Training - experiment/eval & Persist - Learnt feature/best model, and (3) Model Deployment – turn prediction into useable outcomes, consisting of Inference, Monitoring - Feature/model drift and Optimize - Cost/Scale. 3. Partner in enabling AI coding productivity for Engineers, translating to ROIs with Finance team. Kiran is also responsible for managing and continuing to lead (1) Chewy Canada Engineering which Kiran and team built from group up as well as (2) Chewy's Acquisition(M&A) for SmartPakEquine.com which includes integrating new team members and it's technology into Chewy.

 Directly managed and lead engineering efforts for below teams 1) Lead ground-up build/launch of Chewy's 1st International Canada tech team with TAM of $7B with ~190 services/57 databases. 2) Lead creation of total of 5 SPAs/Microfrontends ground-up on Path to Purchase at Chewy, of which 2 SPAs were Home & Search SPA on B2B eCommerce site above and rest 3 were Cart, Checkout & TY Page on Chewy.com that helps support $9B friction free conversion funnel Avg Onload times for Cart page went from ~4s to ~2.3s, Checkout went from ~5.4 to ~2s, TY page went from ~59s to ~1.7s leading to anticipated $200M YOY incremental revenue.

1) Lead ground-up building of B2B eCommerce site in 9 months with an estimated TAM of $xB. Home & Search SPAs were built from scratch with new brand/site experience, segmented search/browse, additions to catalog, & differentiated pricing with a team hired from scratch. Lead team through cross-team deliverables (12 teams) & architecture creating an extensible framework of multi-site eco-systems reducing launch time for 4 other new programs 2) Lead 2 data-migrations based Microservice creation for 2 (User & Cart) most called services called on every Chewy.com page load. Creation of User data microservice, migrating ~100m records for ~18m Customers with zero downtime from PostGres to RDS & ElasticSearch with performance under 12ms for ~5k RPS for p95 with 6x scale support and migration of Cart service from RDS to DynamoDB (NoSQL) tested for 30x scale support also with zero downtime mitigating risks for AWS regional outages. 3) Started Chaos Engineering at Chewy & Production Performance Testing tool, used by all engineering teams for peak preparation. 4) Lead Reviews Platform (2MM) Product reviews. 5) Created Yelp like pages for Businesses. 6) Measured team engagement by weekly pulse survey, improved it by ~25-40%. 7) Lead Senior Product Manager through defining PRFAQs, BRD for 2 new forward-facing initiatives (Pet adopter) & Outbound sales tool.

 Summary : Pitched, Product Managed & built Gamification Platform and directly managed & Lead Software Engineering teams (1) In-store eCommerce website (~$350m revenue/year) , (2) Catalog, (3) Pricing, (4) Promotion Analytics, (5) Recommendations, (6) Android teams (~$90m revenue/year): Lead tech strategy by Quarterly/Yearly planning. Re-structured team for faster delivery as well as talent development.  Deliverables included: •Lead & Architect – Instore eCommerce website which has 10million orders/year & $350 million incremental revenue/year using robust resilient microservices based architecture based on Spring Boot with high QoS and 26 Web Services Integrations including French Canada. • Delivered ML based – “Recommendation Engine” with Data scientists & Software Engineers. • Lead & Delivered Promotion Analytics Tool used to find incremental value of promos based on Seasonality, Geography, Marketing Vehicle & Dept using ML Algorithms like Timeseries(ARIMA, STL), Linear(Elastic-net,glmnet,Lasso/Ridge), Unsupervised like PCA/K-Means. • Delivered 37m personalized recommendations & coupons on In-store receipts (Xerox). • Pitched “Ops Gamification Platform” idea to leadership which was approved. Lead engineering & Product Management effort to implement ML based solution that improved Selling/Manufacturing Ops metrics from 65% compliance to 95+%. • Managed delivery of Pricing engine with self-serve publishing to under 30mins with peak of ~100 RPS. • Manage Catalog Mgmt tool to improve Self-service Publishing of new Products within 24 hours window. • Staffed a QA Automation team that reduced QA from 10-person days to 12 mins in Pricing Engine. • Manage – Development of Android app on Xerox Self-serve printers used within Stores with ~$90m revenue.

 (1) Lead & Architect for Instore eCommerce website (without Payments as it was POS) for Customized Print Orders that generated $350 million of incremental revenue per year and 10k orders per hour. (2) Built ML based – “Recommendation Engine” with Data scientists & Software Engineers for personalized recommendations. (3) Lead migration of above Order capture platform on-prem services to Internal cloud platform.  Development Tools: ● Machine learning - Python & R, RMarkdown ●Application using Java, JDK 8, Spring Boot, Spring Cloud, Spring Batch, Hibernate/JPA, Oracle , SQL Developer, Couchbase (NoSQL), Hadoop, Hive, Kafka, Rabbit MQ. ●Cloud Deployment/Logging Tools: Azure/AWS, Git, Jenkins-CICD Pipelines, ●Performance/Automation Test tools – Jmeter, Splunk, Selenium,Appium

 Lead & Architected engineering efforts for below 2 teams. They were (1) Within “System-One application” which generated $1Billion revenue per year built modules used to facilitate various claim operations like Life Claim, Leave Claims, Short Term disability, Long Term Disability (2) Built “Automated Claim correspondence” Application that helps Liberty Mutual Claim managers with a self-service interfact to generated correspondence (US Postal, Fax & Email) for their Group Insurance Clients.  Deliverables included: • Lead & Architect – Application that provides claim managers self-service interface to generate auto-mated correspondence (US Postal, Fax and email) for claims for their Group Insurance Clients. Improved delivery by getting a 1-month delayed project & Earned Trust using Agile Sprints using Microsoft Project Planner (MPP)- Burn down charts. • Optimized Return Address Module with Algorithms from O (n2) to O(log n) improving 25% overall performance. • Optimized Spring + Drools life cycle to reduce time from 2000ms to 300ms - approx. 85% performance improvement Lead & Architect Modules within - “System-One application” which was used to facilitate various claim operations like Life Claim, Leave Claims, Short Term disability, Long Term Disability which generated $1Billion revenue per year. • Improved performance for a non-functional requirement of System-One app which needed tuning by merging of 45+ projects to set of 8 projects dividing them into Java layers (viz. Base -> DAO-> Batch & Base -> DAO -> Utilities ->EJB -> Web) lead to improvement of 66% performance improvement in HP Load Runner Performance Tool simulated for 1000 concurrent users • Redesigned multi-level batch jobs to use Spring Batch Framework with better restarting & resiliency. • Conducted various training sessions on Spring, Design Patterns for co-workers.

 Lead & Architected modules for building a strategic Java based Self Service data-warehouse product development which was a “Platform as a Service” (PaaS) and “Software as a Service” (Saas). This product was used by Investment Bankers & Investment Managers to generate highly sophisticated and customized report which used to help them make their investment decisions which could be a function of various parameters like portfolios, stocks, Investment Rating agencies etc. • Developed Job Controller Framework for highly decoupled, high-resilient Load Balancing Application using JMS–ActiveMQ. • Data-Model Design Module gave user self-service interface to design, develop & persist XML based meta-model structure of java-based Data-Model / Categories into database which enabled multi-user editing of Data Models • Developed Java based Asynchronous DataMart Refresh for hierarchy of Data-Model to do a series of OLAP (Online Analytical Processing) calculations like Sum, Join, Transpose etc. triggered by Job Controller Framework. • Developed a Configurable extract creation Framework + Application to send OLAP data files in sftp in csv, xml, xls & txt formats to multiple consumers at various file locations.

 Developed Full stack workflow between UI screens to DB using Struts-Hibernate in “Incremental offers” application for Bank of America – which displayed offers at ATMs for users of Bank of America, Also enabled Bank Admins with a custom approval hierarchy such that Admin would be able to only Add, Modify or delete but for the offer to get displayed it needed other Sr.Admin to approve.  Developed various Full stack workflow between UI screens to DB using Spring-Hibernate on developing “Goods in Transit” application for Honda which enabled Honda Car dealers capability to predict their requirement of different kinds of Honda cars up to 9 months in advance. Objective was to streamline the production of cars, reducing the excess inventory that could be sitting in their warehouses

One of the Top 10 Engineering schools in India!

MS Candidate with specialization in Financial Econometrics & Financial Valuation!

Mumbai University