Mar 2021 – Jun 2025 · 4 yrs 3 mos · Toronto, Ontario, Canada

• Led development of real-time inventory synchronization integration between merchants'
• Buy with Prime and Amazon FBA accounts. Coordinated cross-functional teams, including
• managing 3 engineers, to deliver the project on a tight deadline. Designed a push-based
• approach to replace the existing pull-based syncing, reducing overhead and improving
• scalability. Deployed the solution, enabling a seamless omnichannel experience for
• merchants.
• As lead engineer, implemented asynchronous refund workflows for enterprise merchants
• on the customer support agent UX. Collaborated with program managers to understand
• requirements, provide critical feedback, and enable successful delivery. This streamlined the
• refund process and improved the customer support experience, while maintaining the
• existing synchronous flows for other integrations.
• Redesigned the deployment pipeline architecture, splitting a monolithic pipeline into
• individual, application-specific pipelines. This reduced manual interventions by 20x,
• improving development velocity and efficiency. Leveraged infrastructure as code to
• automate pipeline setup, standardizing the process for rapid onboarding of new projects.
• Spearheaded the migration of a legacy warehouse communication service onto a new
• Fulfillment architecture as part of an organization-wide initiative. Collaborated
• cross-functionally to define data contracts, contribute to system designs, and prioritize
• critical infrastructure requirements. Delivered the project within a tight timeline by breaking
• down the work into phased implementation. Ensured a smooth transition without
• disrupting active warehouse processes and avoided any production issues. This successful
• migration paved the way for onboarding the new architecture to other supporting systems
• in the warehouse ecosystem

Jul 2020 – Dec 2020 · 5 mos

• https://tech.oyorooms.com/how-we-rank-hotels-at-oyo-416e8363a1df

Jun 2018 – Jun 2020 · 2 yrs

• Rule Based Ranking system:
• Single handedly designed and developed a linear rule based system, which would rank the hotels in a users search listing. The system was configurable enough to handle various business use cases easily.
• The service currently runs in the Tier1 Search page of the OYO app, and is currently serving around 15000 TPM.
• Technologies used:- Java, Redis, Kafka
• Intelligent Ranking system:
• Deep dived, designed and lead the team of 6 people through the development of a system which can launch ML based ranking models on location, search type etc. basis
• The system is such that it allows developers to run there models on a subset of production traffic and role it up from there
• The system has been developed in multiple layers, each layer tries to optimise for a business metric.
• Developed some initial models based on XG-Boost and Lambda rank, to launch in high traffic cities. We were able to improve the booking percentage by around 10%.
• Technologies used:- Java, Redis, Kafka, AWS Sage Maker, Spark, AWS Lambda, Python
• Clickstream Events management service:
• Designed and lead the team through the development of a system, which stores users clickstream events in, near real-time and provides aggregations as requested by the clients.
• This service is essential for showing relevant search results to the user while he/she is still in the same session.
• Technologies used:- Java, Redis, Kafka

Dec 2016 – Jan 2018 · 1 yr 1 mo · Gurgaon, India

• Carrier Authority:
• Designed and developed the flow for onboarding carriers to the new carrier authority system. Developed the end- to-end flow for storing fetching and updating the contact details of the carriers.
• Developed a fail-safe strategy to migrate the carriers from the legacy system to the new system and backfilled the data for the time all the clients were not on boarded.
• Technologies used:- Java, Mithril, DynamoDb
• Carrier Financials:
• Worked with the team to design the overall architecture for carrier financial systems, and defining the roles and responsibilities of each component and their interactions among them and with upstream and downstream systems.
• Created an engine to output the invoices in multiple formats(csv, pdf, xlsx) based on client requirement. This expected templates and the data objects and returned the invoice.
• Designed, and developed the flow for auditing invoices information and creating backups. The system was deigned in a way that it would not interfere with the production flow. Was done using AWS tools like DynamoDbStreams, AWS Lambda, AWS kinesis firehose, and stored the data in S3.
• Technologies used:- Java,Apache FOP, Apache POI, Apache Velocity, AWS SWF, DynamoDb, AWS lambda, Kinesis Firehose, S3, SQS.

Apr 2015 – Nov 2016 · 1 yr 7 mos · Noida Area, India

• Order Management System:
• Designed the end to end flow for Orders coming on the app. Developed major parts of the system, which includes, Cart Creation, Order Creation, Assigning it to the respec- tive seller etc.
• The project also involved processing of the order on the seller side, handling discounts for Exchange Products.
• The system is robust and scalable and handles more than 1000 new orders each day.
• Technologies used:- Celery, SQL, Django, Python, Redis, Ansible
• Delivery Management System:
• Designed the end to end flow for Orders coming on the app. Developed major parts of the system, which includes, Cart Creation, Order Creation, Assigning it to the respec- tive seller etc.
• The project also involved processing of the order on the seller side, handling discounts for Exchange Products.
• The system is robust and scalable and handles more than 1000 new orders each day.
• Technologies used:- Celery, SQL, Django, Python, Redis, Ansible
• Payment Reconciliation System
• Designed this internal system for reconciling the payments from different payment gateways and COD from delivery partners.
• Fetched data from multiple systems, processed them to generate a report daily for keeping track of the money, inflow and outflow, from delivery partners and the sellers.
• Replacement and Returns Flow
• The system handled replacement and return requests from the customers.
• The request is first validated and then a reverse flow for delivery partners is run.
• Bajaj Finserv Integration
• Integrated the Bajaj Finserv payment gateway to the app.
• This gateway was different from other payment gateways in a way as it had EMI's, the gateway had multiple rules without which it could not have been activated, we had to sell warranty products as well along with the product, had to send a proof of delivery to Bajaj, etc.
• This involved major changes in various systems, including orders, payments, warranty, etc.

Sep 2014 – Apr 2015 · 7 mos · Noida Area, India

• Dived deep into the complexities of Redis Server to test the worst case performance and functionality of a new search feature for filtering out the details of various advertisements to be displayed on the video player.
• The feature made a huge impact in terms of customer satisfaction by reducing the time for the process from hours to milliseconds.
• Found a rare security bug, thereby making the system reliable and secure. It required rigorous testing across various platforms to find the root cause.
• Automated testing for the process of advertisement ingestion thereby reducing the testing time for the application from 2-3 days to 20 minutes.
• Worked on a variety of environments, which run on Redis Server, Linux, Python, PHP etc.

Feb 2014 – Aug 2014 · 6 mos · Bangalore, India

• Migrated the existing customer data to the new buy-flow thereby reducing the page rendering time by a few seconds.
• Learnt the basics of the ATG Architecture and resolved some existing bugs in the released versions of AT&T Partner Exchange website.