Nextdoor Ads Center and Events Tracking System: Key contributor to the self-service Ads platform enabling businesses to create ads. Developed new GraphQL and REST endpoints, Kafka events, and React frontend for key features like Ads subscriptions, SSO SAML assertion between different services for user convenience and secure login/signup, and created pipelines sending metrics to external platforms like Google Analytics, Facebook CAPI, and Pixel.

• Internal Tools Development: Played a central role in the development and maintenance of essential internal tools, catering to the needs of case support agents in direct customer interactions. Key contributions include creating tools such as Postcard Invitation, Bulk and Individual Content Removal and Reinstatement, Activity Logger, Content Moderation, and User Permissioning.

• NPS Survey System Architecture: Spearheaded the design and implementation of an NPS survey system, orchestrating the distribution of surveys to 25k users weekly and capturing their responses. This system significantly aided the research marketing team in tracking repeat customer rates, identifying trends, and analyzing customer loyalty patterns across 13 countries, enhancing Nextdoor's product strategies.

• Performance Optimizations: Optimized performance of several pages by identifying and refactoring 30 slow API endpoints, by rewriting the logic, rewriting slow PostgreSQL queries, precomputing materialized views, migrating high-volume queries to Snowflake, and implementing Redis caching.

• Third-Party Signups: Integrated third-party signups with Facebook, Apple, and Google Single Sign-on.

• Log Monitoring and Alerting System: Designed a log monitoring and alerting system, enabling services to post logs into AWS CloudWatch and Datadog, resulting in reduced incident reaction time from hours to seconds.

• Model Evaluation Platform: Led full-stack development of a model evaluation platform, which provided scoring like Precision, Recall, F1, AUC, K score of machine learning models, a niche solution implemented after reading some research papers. This feature contributed to a 3% increase in customer retention and generating ~$1-2M in incremental revenue/year.

• Active Learning Tool: Built an in-house data annotation tool to automate labeling of 10K records and train ML model using a small dataset by creating an Active Learning tool, saving 4-5 hours of analysts' time/week.

• Computer Vision Platform: Designed and implemented a image recognition system using external API's like Brandwatch and Clarifai. This feature allowed uploading of images into the web app and displayed relevant annotations on the image along with their percentage scores given by the REST API.