Resume

Senior Software Engineer - Feature Management & Experimentation

May 2024 - Present

Projects

• Led implementation of Warehouse Native Experimentation support, enabling customers to run experiments directly against their data warehouses
• Architected the usage reporting data pipeline, including the Temporal Workflow and dedicated Tinybird (ClickHouse) MVs, to power Harness's flexible pricing model for FME
• Engineered the MCP server toolset for Harness FME, enabling AI agents to autonomously identify and clean up stale feature flags and analyze which code paths are live in production

Leadership

• Serve as a Tech Lead on a 15-person engineering team, owning architectural decisions and coordinating cross-functionally with product and stakeholders
• Facilitate Agentic AI trainings for the 80+ engineers in the FME module to build team knowledge of the latest innovations, increasing AI adoption rate and creation of internal marketplace for Claude Code plugins

Software Engineer - Measurement and Experimentation

Jan 2023 - May 2024

• Redesigned experimentation from a feature flag add-on to its own dedicated entity and services, establishing the foundation for all modern experimentation calculations
• Migrated experimentation platform from Databricks Spark to Tinybird OLAP data platform, enabling real-time data analytics on experiment results

Software Engineer - Office of the CTO

Jul 2021 - Jan 2023

• Implemented a RAG-based chatbot on Split's documentation and blogs to help customers navigate Split's product
• Prototyped a system for ingesting external feature flags into Split, validated technical feasibility and customer demand, and handed off to Engineering to productionize
• Developed various OpenFeature providers to allow vendor-agnostic use of Split's SDKs, including Java, Go, JavaScript, and Python providers

Software Engineer

Jul 2020 - Jul 2021

• Built and shipped a full-stack Notification Center microservice for Webex Control Hub, owning both backend APIs and frontend integration
• Designed and compared two approaches to detecting Webex call quality degradation: an autoencoder-based ML model and a statistical threshold detector using 95th-percentile sliding window analysis