The Situation: At NVIDIA, I was responsible for shipping server-class hardware designs for autonomous vehicle (AV) platforms. This role sat at the critical junction of high-performance computing and automotive-grade reliability, where I owned the 3D modeling lifecycle and the formal Engineering Change Order (ECO) submission cycles.

Why it was difficult: AV platforms operate in extreme environments where mechanical drift is not an option. The difficulty lay in the forensic tolerances required for server-class hardware. Any interference or thermal variance could compromise the sensor-processing unit. I had to manage the complex ECO cycles while ensuring that new component layouts didn't introduce systemic risks during high-speed validation.

How I worked through it: I applied a systems-led governance approach to the design and validation pipeline:

• Forensic Design Audits: I conducted exhaustive DFM (Design for Manufacturing), tolerance, and interference checks. By treating the 3D model as the source of truth, I ensured that every layout was physically optimized for mass production before it ever hit the factory floor.

• Cross-Functional Synthesis: I partnered directly with systems engineering to accelerate the validation of new layouts. This ensured that the mechanical constraints were always aligned with the electronic requirements of the autonomous modules.

The Outcome: I successfully delivered server-class hardware that met the rigorous safety and performance standards of the AV industry. I am proud of this because I proved that infrastructure rigor, the ability to reduce iteration time while increasing design fidelity, is the only way to scale complex autonomous systems. I moved the needle from experimental prototype to production-ready infrastructure.

During an 8-month residency at Google X, I operated at the intersection of hardware prototyping and operational reliability. I was tasked with architecting the physical framework that defined the robot's perception of the world.

Why it was difficult: The primary challenge was sensor drift. In a moonshot environment, even minor mechanical variances can compromise the integrity of the entire autonomy stack. I had to ensure that every robot in the fleet possessed an identical physical baseline, despite the complexities of outsourced manufacturing and rapid prototyping cycles.

How I worked through it: I acted as the strategic liaison between design and deployment, enforcing a forensic baseline across the hardware lifecycle:

• Mechanical Governance: I engineered a suite of precision mounts with a strict 15° pitch baseline. By standardizing these mounting surfaces, I ensured that sensor data remained high-fidelity and repeatable across different environments.

• Operational Standardization: I developed and kitted comprehensive installation playbooks. This moved the deployment process from custom assembly to a standardized predictable reflex, ensuring 100% environment parity across the fleet.

• Infrastructure Management: I owned the quote-to-part pipeline, auditing external vendors to eliminate supply-chain bottlenecks and performing low-level debugging/flashing of motor drivers to maintain the platform's "operational heartbeat."

The Outcome: I successfully eliminated mechanical sensor drift as a variable in the autonomy stack, providing a stable architecture for the engineering team. I am proud of this because it demonstrated my ability to translate complex process analyses into actionable narratives for engineering leadership and proved that physical governance is the bedrock of reliable software performance.

(November 2019) "Women In Engineering", Design World magazine (Cleveland, Ohio)

In summer of 2016 I wrote an article talking about Diversity in Machine learning and the real effects that a lack of diversity can have on product experience. Later that year I received a request from Capital one to appear at their Humanity.ai confernece in San Francisco and deliver a keynote address based on the article. Within that room was the conference organizer for the 2017 National Non-Profit Technology Conference which is was also invited to and delivered my first speech in Washington DC to 2000+ people. I also updated the article in 2017 titled, Recognizing Cultural Bias in AI.

Video from BBC

BBC launched a 100 women campaign that highlighted 100 women leaders around the world. My colleague Roya, who is a product designer, was featured as one of these women and put together a team that included myself, an industrial engineer and another product designer on the Forbes 30 Under 30 list. 

BBC came to the San Francisco area in order to film a week long hackathon where we were to build a product that would help women in Silicon Valley break the glass ceiling. 

Our first idea, was an early iteration of the idea I am using in Entrepreneurship 415, a wearable that would encourage women when they were in high pressure and performance events. 

The second idea was a world map painted on a large wooden structure that represented the global location of women at work across the world. The interactive display allowed audience members to go up and play the stories of women being sidelined. 

My work was on the engineering side to make a prototype of the backend. 

(April 2020) "Idaho's Formidable Women", Idahome magazine (Boise, Idaho). 

https://issuu.com/idahome/docs/v2_i4_for_issuu_-_single_pages/26 (accessed March 31, 2020)

With my colleague Alivia, we launched a podcast that has featured 3 conversations around education and technology. With Alivia working as a Data Scientist and myself as a Robotics engineer, our conversations lend towards advanced technology with a societal lens.

In our episodes we aim to have a conversation with a guest. So far we have hosted an AI expert and a behavioral health professional. 

In this undergraduate research position, I developed multiple manufacturing process for ceramics, using Solidworks and various laboratory equipment. The end result was a variety of projects in micro-fluidic pumps, space antennae and thermoelectric generators. We presented this research at the Boise State Undergraduate Research Conference Spring 2014.

Abstract:

Thermoelectric generators (TEG) directly convert thermal energy into electricity without any moving parts. Currently the TE module package accounts for a significant portion of the total cost. Our research is investigating the use of LTCC with silver vias and traces as an innovative packaging solution. An LTCC package has been designed and fabricated to demonstrate this approach. TE elements are bonded to the LTCC using copper traces that also form electric connections between the elements. The device has a maximum operation temperature of 600°C. Finite element modeling was employed to optimize the design to achieve minimum thermal resistance through package and to maximize heat flow into the TE elements. Preliminary simulation results predict that the LTCC packaging can significantly enhance device performance when compared to the traditional package. LTCC packaged TE module prototypes were also fabricated. Four substrate patterns will be described along with the modeled and experimental thermal resistances of each design. Our ongoing work is focused on optimizing both the package design and the fabrication process in order to achieve the highest performance and reliability. The cost reduction in TE module packaging will open a great deal of opportunities in waste heat recovery applications using cost-effective thermoelectric generators.