At Verkada, we build security cameras that detect action, identify danger, and help keep people and places safe and secure. Software engineer in search and computer vision team.

Reduced face search indexing latency from days and hours to minutes by implementing a global cache system

Reduced latency of the person of interest notification pipeline by 84% through a single entry batch instead of multiple batching steps

Sharded the face search service for scalability

Developed a debug platform to save log entries inside a database across the Verkada BE

Technologies used: React, Django, Kubernetes, MySQL, Redis all powered by AWS

Troubadour is a text-to-speech chrome extension with a current user growth rate of 2x every 3 months

Designed and Built a chrome extension with vanilla JavaScript which reads aloud web content using text-to-speech. Reduced load time and improved user experience compared to competitors

Designed and Built a dynamic website which communicates with the chrome extension through an API

Configurable to user preferences including over 40 languages, playback speed and others

Operating system and browser voices are available for free, premium voices generated with the latest AI tech-nology available for purchase

Technologies used: React, Django, Kubernetes, Wordpress, MySQL all powered by Google Cloud and continuously integrated with Google Cloud Build pulled from Github

Lead and collaborated on teams. 7 research projects. 620+ citations. Scientifically advised by Prof. Federico Capasso, Prof. David Weitz, Prof. Don Ingber

Projects (commercialization effort):

• Designed and Built a meta-surface doublet in the visible through building a ray tracing engine. (MetaLenz Inc.)

• Designed meta-holograms w/ Gerchberg-Saxton algorithm. (None)

• Developed a simulation framework for multi-layer meta-surfaces. (MetaLenz Inc.)

• Developed a sequentially-stacked microfluidic droplet maker. Publication pending. (Tune Bio Inc.)