Oct 2021 – Aug 2024 · 2 yrs 10 mos

• Building scalable and extensible next generation retail and planning software suite using cutting edge technologies.

Apr 2019 – Sep 2021 · 2 yrs 5 mos

• Conceptualised and Drove Retail as a Service which is a zero touch and cloud based retail solution which can be leveraged across Flipkart and its sister companies.
• Grew the team from 40 to 130 over the period of 3 years.
• Team contributed significantly to make Grocery, Express Delivery and FnV happen for FK.
• Team made the first intelligent and smart automated buying and placement of inventory happen for FK.
• Team made the exchange construct happen for Flipkart to drive more customer affordability and better value recovery from the used products.

Apr 2017 – Mar 2019 · 1 yr 11 mos

• Built the engineering team for Project ICE (code name). Project ICE is one of the new exciting private brand initiatives under development at Flipkart. Its mission is to provide low income consumers access to products which are of top quality comparable to top national brands. Eventually, the vision is for this brand to carry products across most categories of Flipkart. Flipsters in ICE work on problems such as advanced information­ retrieval algorithms, massive scalability,predictive analytics, machine learning, NLP and large scale applications that enrich the user experience, to name a few.
• Working on a challenging problem of extracting actionable insights from customer reviews and catalog images for new product development. As well as automating complete business process for private labels.

Jul 2016 – Mar 2017 · 8 mos

• Drove Business Unit wide exercises and beyond such as,
• The application changes needed for infrastructure cloudification exercise across the across 30+ applications using variety of tech stacks which enabled infra team to make Machine Allocation Self Serve and optimal utilisation of Hardware resources.
• Across Ecosystem level changes to provide early visibility into orders/fulfillments for better Ekart asset utilizations.
• Externalization of Fulfilment Offerings to provide a single platform for Fulfilment with a key capability to aggregate supply and aggregate demand to clients who ultimately use this platform as their only fulfilment platform.
• Multi Package Shipments to relaunch furniture with a goal to scale this category significantly.
• Was part of defining and evangelising the competency framework across organisation for PMS.
• Team delivered some of the business critical projects such as MPS, Groceries, CAD v2 etc during this period as well as ensuring smooth execution of BBD 2016!
• Ideated , Brainstormed and Kickstarted entire orchestrator system rewrite exercise.
• Also had got the additional responsibility of managing Ekart Specific orchestrator systems. Total team size had grown to 30+.

Jan 2015 – Jun 2016 · 1 yr 5 mos

• Lead and Managed Physical, Returns , Digital and After Sales Service fulfillment teams of Flipkart whose main objective is to come up and execute on the fulfillment plan by orchestrating with various downstream and upstream systems so that original promise made to the customer is kept at any cost.
• Team Size: 18+
• Team delivered some of the business critical projects such as Dropship ,PreOrders ,Jeeves integration, Gateway , New Subscription Fulfillment , Ship Together , Ship Separate for VNFs , Fulfillment Dashboard Enhancements, Surface Tier Support, D + I support, Capacity based forward orchestration, Smart Pickups, O2I visibility etc
• Team delivered some of the reusable services such as Anusuchak(Scheduler as a Service), Mercury(Scalable Non transactional data ingestion service) etc
• Team delivered some of the simplification/re architecture projects such as Returns Orchestration Separation, B2B/B2C separation etc.
• Team also delivered ML Based range SLA prediction for providing range of promise to the customers to improve the order conversion. As part of this team delivered E2E reusable ML pipeline for data aggregation,splitting ,mapping, filtering ,training and prediction.
• Team also excelled in operation excellence by reducing significant oncall load through right automation and systemic long term fixes.
• Ensured the scaling of the systems for Big Billion Days.
• Ideated and Drove the efforts towards making After Sales Service and Returns Promise Live.
• Was part of the tiger team for arriving at Blueprint for fulfillment ecosystem to take the fulfillment experience of the customers to the next level.

Oct 2013 – Dec 2014 · 1 yr 2 mos

• Lead Cross Cut Digital Santa Integration project which gave tremendous power to support offers on digital items. This considerably increased the sales on EBooks.
• Built the Scalable and Extensible Digital fulfillment system which covered the fulfillments for EBooks, EGV and Elearning.Built the entire system from scratch with the team size of 4. (Including me).
• Launched the ELearning vertical from scratch to fulfill the digital codes such as exam codes which had a high expected topline impact.
• Built the Scalable and Extensible After Sales fulfillment system which covered the fulfillments for after sales services such as installation , uninstallation, technician visit etc for the large products. Also built the service desk to orchestrate the fulfillment with external brands.
• Built the entire system from scratch with the team size of 3. (Including me)
• Tuned Elastic Search data store of Dexter ,the reseller fraud detection engine which reduced over all average response time for fraud score computation from 15+ seconds to 300ms.
• Architected and designed scalable digital promise engine which provided availability and serviceability checks for digital products such as EBooks, Elearning and EGV. Digital PE was benchmarked with throughput of 800rps per medium sized VM at the
• concurrency of 5 with 95th percentile latency 7ms.
• Initial Member of designing and developing an open source Aesop system which is a Highly Available, Scalable and Low Latency change propagation system which can processes 50,000 events per second (Capable of larger numbers based on hardware and use case).Change Propagation happens in real time. Changes are propagated across data stores in 100 ms in best case.Based on log mining approach. Processes binlog of MySql or WAL of HBase as source. Events are then processed, stored in Relay. This was inspired by LinkedIn's Databus project.
• More Details: https://github.com/Flipkart/aesop

May 2012 – Sep 2013 · 1 yr 4 mos

• Joined Digital Platform team of Flipkart whose main vision was to build a scalable digital platform which can retrieve, process, store and deliver multiple digital contents like ebooks, mp3s, videos etc starting with EBooks launch.
• Built,Owned and Maintained Digital Content Delivery modules to retrieve digital content from storage after right set of validations on the incoming download requests and deliver the digital content in a secure and efficient way with minimal latency for EBooks and MP3s.
• Built a scalable crawler to crawl the required competitive intelligence from competitor sites for required products. Extended the same for the specific EBooks as well.
• Built a light weight , scalable and performant location service which provides location details such as Country Name, Country Code, Region Name, City Name and Proxy Check for the given IP. Supported throughput of 50k RPS and was used across Flipkart at a web scale for fetching location details.
• Owned and Worked on various EBook consoles such as ,
• Supplier Console: Register and Configure Ebooks Supplier metadata
• CMS Console: View/Edit Product metadata
• CS Console: Check Download History, Increment Download count

Jan 2010 – May 2012 · 2 yrs 4 mos

• Cisco Active Network Abstraction (ANA) provides service providers and other carrier-grade network operators with a comprehensive, network management solution for IP next-generation networks. ANA creates and maintains an accurate abstract model of the network in near real-time to manage network elements, with support for all major Cisco service provider architectures and can integrate with network management systems (NMSs/OSSs).Virtual
• Network Elements (VNEs) are the Network devices such as routers/switches managed by ANA. The VNE subsystem is the interface to the network. Its main role is to discover and maintain an accurate abstract model of the network, via device queries and event notifications.
• Lead the solve(EVNE) along with other 6 engineers to update the network management view in near real time based on the configuration change event vs periodic poll event.
• Received multiple Cisco CAP awards in this period.

Aug 2006 – Jan 2010 · 3 yrs 5 mos

• NBI aims at providing a North Bound Interface between an NMS/OSS and the Huawei Element Management System. NBI provides a 3GPP compliant north bound interface which any 3GPP compliant NMS/OSS can use to manage the Huawei network elements. Since usually the telecom network consists of equipments from different vendors, considerable amount of effort goes in integrating the equipments from different vendors. If we have a standardized North Bound Interface (NBI) developed, this can save the considerable amount of time and money in integrating the equipments from different vendors. NBI is the back bone for strong Network Management Network.
• I had almost independently developed this version of NBI. The Major features developed in this version were Fault/Performance/Configuration/Subscription Management Operations and notifications.
• Got Spot Award for the best quality contributions.