• Engineered fleet monitoring dashboard system using Grafana and InfluxDB to analyze camera, GPS, and sensor

telemetry data, enabling continuous monitoring of selected fleets post-software deployment and ensuring rapid

detection of issues during frequent update cycles

• Developed weighted scoring system that analyzes camera hardware error registers, data quality metrics over

time, creating Python-based visualization tools that enabled data-driven maintenance decisions

• Built extensive Python automation framework with JIRA integration that provides tooling for debugging and mitigating real-time device issues, collaborating with field and device teams to reduce resolution time

• Led the development and maintenance of a full-stack software system, integrating C++ Arduino

based code across various robot types and building Python infrastructure for end-user

applications, with a focus on robust software architecture designed to scale and accommodate

multiple variations of robots

• Architected and implemented a full-stack LTE communication infrastructure, enabling seamless

two-way trans-mission of encoded message packets and improving connectivity for robotic

systems using Cellular XBee devices and integrating with AWS IoT Core, Lambda, DynamoDB, and

SMS notifications

• Optimized camera-based situational awareness systems by leveraging IMU sensor readings for

horizontal image correction. This included static cameras placed around the sphere of the payload

and a rotating camera on a pan-tilt system, significantly enhancing situational awareness as

requested by customers

• Implemented an OS-independent video receiver featuring auto-detection via FFMPEG and

GStreamer, using threading for flawless device connection management

• Shipped a major software release affecting numerous customers, performing extensive testing

across various hardware revisions to ensure high reliability and performance

• Mentored numerous interns and UC Berkeley research students with a focus on novel research and

publication on machine learning/computer vision methane detection from multimodal sensor

readings

• Collaborated directly with a small in-house manufacturing team for provisioning and deploying

robotic systems, enabling fast iterations on revisions

• Implemented a Machine Learning Pipeline to automate defect detection on printed circuit boards

• Performed image classification and object recognition with OpenCV and TensorFlow

• Performed Agile Software Development with Jira and GitLab

• Developed software for SuperTech FT’s robotics products (RETA, an intelligent web-based app for

remotely teaching, learning, programming, and controlling robotic arms)

• Trained and tested autonomous navigation using TurtleBot with LIDAR in ROS

• Created and taught robotics and machine learning education to students at San Francisco’s Code

Tenderloin