• Developed LLM orchestration and filtering systems that reduced token usage by ~60–70% on large queries, lowering infrastructure costs and unblocking responses for enterprise customers who were previously timing out at scale.

• Built a continuous DevEx feedback pipeline that automatically surfaces weekly engineering insights within existing developer workflows.

• Launched the Activity Recap page, surfacing recent engineering activity with AI-suggested discussion prompts to improve 1:1 conversation quality.

• Migrated cost capitalization reporting from Google Sheets into a native in-app experience, improving reliability and reducing manual overhead.

• Delivered full-stack self-service automated cost capitalization, enabling customers to define custom rules for automatically flagging capitalized projects and directly supporting tax benefit workflows at scale.

• Co-developed and launched a new cognitive science course, scaling enrollment from 100 to 700+ students over four years.

• Led and managed a team of 17 teaching assistants, overseeing grading workflows, training, and instructional quality.

• Designed and iterated on course assignments and rubrics to improve clarity, rigor, and student outcomes.

• Built scalable operational systems for large-enrollment courses, including TA coordination, feedback pipelines, and student support processes.

• Redesigned JSON configuration files to support future scalability, which allows for up to 100 active nodes on the network.

• Designed an API to transmit voice and text data via a third-party modem for a wireless radio communication system.

• Employed best practices such as writing unit tests, regression testing, and refactoring old code to utilize C++17 functions.

NextGen Battery Embedded Software Engineer

• Architected and wrote embedded Linux/RTOS applications for a battery management system that monitors state of charge and state of health for eVTOL propulsion systems.

• Developed C++ drivers for peripherals on battery management computers that communicate via SPI and CAN.

• Wrote VHDL code for MAX V CPLD that interfaced between an STM32F7 processor and custom controller board.

• Designed a high power test system to simulate conditions for 4 vehicle mission profiles.

Counter-UAS Avionics Team Lead

• Managed a team of 5 engineers and 1 intern to iteratively build and prototype a computer vision-based drone in 1 year.

• Created the avionics manufacturing package and conducted a trial run for building a batch of 8 drones.

• Recognized by Aurora Flight Sciences as the inaugural “Team of the Quarter” and by Boeing for “Team of the Month”.