Shivangi

Personalization Team

• Currently maintain a legacy recommendation serving system. Scaled the system the from 1k rpm - 200k rpm to support growing number of customers on the platform
• Lead the design and development of the new recommendation serving system in Go.This rewrite brought down the infra cost as we were able to serve more traffic with less pods and reduced our avg latencies from over 150 ms to 10-20 ms
• Built an in-house user surveys system in go for post order feedback and KYC to serve better recommendations
• Built a real time model inference service in python 3, to serve real time user level recommendations
• Currently oversee the scalability and reliability of all personalized assets on the blinkit app
• Also helped build the vector/embedding service for improving user experience and a rule engine to accommodate category team inputs for festive season in recommendations

Consumer Tech Team

• Worked on improving the ranking of products and customer experience on product listing pages on the Blinkit app
• Worked in the Diwali Festive squad which built the initial user experience and tech to sell gold and silver coins on Blinkit app
• Helped in design and building the tech and compliance required to Sell Iphones on Blinkit app
• I led a project which would convey to the customer , when a product might next be available or when it was sourced in case of perishables. This project was built when we were still Grofers. Now we have brought back the sourced at feature for perishables in Blinkit.

Supply Tech Team

• Worked on building dropzones and dropzone sortation logic for our warehouses which reduced the time and chaos taken to QC during dispatch significantly
• Built a near real time metering dashboard, which was used by our category and purchase managers to decide when to place orders and what quantity for certain skus. This was crucial , as it helped us plan better for festive demands by leveraging customer purchase signals
• Helped build the tech for multi order picking for pickers in our warehouse, which reduced the cost per order
• Designed and developed the initial POC for aisle based picking, which is currently used in our warehouses for better resource utilization
• Built the initial tech for wave picking in our warehouses to optimize resource utilization in case of bulk stock transfers
• During COVID-19, designed and developed an alternate supply chain model called City warehouse. We ran this on and off during lock down to cater to government restrictions and increasing customer demand to keep the business running.