Dec 2018 – Feb 2021 · 2 yrs 2 mos · Bangalore Urban, Karnataka, India

• Data Platform Team
• Designed and developed the central alerting platform of Swiggy from scratch.
• The platform currently powers over 300 online (realtime) and offline alerts.
• Tech stack included Java, Esper, Spark, Kafka, Redis, MySQL, Qubole, Snowflake, etc.
• Data Science Platform
• Worked on implementing bloom filter to avoid false negative calls to feature store (Dynamo
• DB, Redis, etc.) for fetching the features needed to inference model. This helped us to
• significatly reduce network calls to feature store and huge cost optimization as well (as a
• result of saved Dynamo RCUs).
• Data Science
• Worked on the task of Named-entity recognition to identify dish item names in the chat
• sentences between Swiggy’s support executive and customer.
• Trained various models like CRF, Naive Bayes, Bi-LSTM, GRU, Char CNN, etc.

Jun 2017 – Dec 2018 · 1 yr 6 mos · Bengaluru Area, India

• Operator Application (Backend)
• Designed, developed and revamped major flows of Operator system so that it can be versioned and digitized properly.
• Scaled the application to support high throughput using distributed data pipeline (Apache Kafka) and improving performance using in-memory cache client (Redis).
• Have been a leading contributor in improving operator backend performance by 5X and have
• helped it scale to more than 4X throughput.
• Tech stack includes Java, Servlets, JDBC, Web Services (Restful), Hibernate, Dropwizard, Hystrix, Kafka, Redis, Docker, MySQL, Maven, Git.