R0 Systems
Founder & CEO | Jun 2024 - Present
I grew up around the chemicals used in farming. With R0, I started building a way to replace chlorine on produce wash lines: plasma-activated water, generated on demand from air, water, and electricity. I founded the company in June 2024 and bootstrapped it to six-figure revenue.
I designed the machine from the reactor out: the plasma reactor and microbubble generator, Venturi gas transfer, drive electronics, and an STM32 hard real-time dosing loop in C/C++. This work goes all the way down to the electronics, the water, and what happens when you actually run produce through it.
In validation, the system outperformed chlorine on disinfection without chlorate or perchlorate residue, kept produce fresh three times longer, and cost less to run than the incumbent chemistry. The work includes three peer-reviewed papers and independent national food-safety laboratory studies. Two patent applications are pending.
OpenTrade
Forward Deployed Engineer | Jun 2026 - Sep 2026
At OpenTrade, I was the first machine-learning hire. I owned the path from live market data to grounded LLM-generated trade cards, then the personalized ranking that decided what a user actually saw. The product grew from launch to roughly 150,000 monthly users in two months.
I replaced hand-tuned feed ordering with a learned model using interaction, content, recency, and market-context signals. I also built the checks that caught unsupported claims, wrong numbers, stale context, and broken output formats before a model or prompt change reached users.
I turned impressions and downstream behavior into evaluation and training signals, defined launch hypotheses and guardrails, and built the backend and serverless routes serving the results. The work covered model evaluation, external data providers, testing, deployment, and what happened after people started using the product.
Sensors, Energy, and Automation Laboratory
Research Associate | Mar 2025 - Nov 2025
At UW's Sensors, Energy, and Automation Laboratory, I worked on the difficult bit between a physical signal and a measurement you can trust. One project was a real-time PPG drowsiness wearable. I built its analog front end, fourth-order Butterworth filtering, 100 Hz 12-bit acquisition, and adaptive noise cancellation that rejected 95% of motion artifact.
Another was a PZT-array acoustic-emission system for U.S. Navy hull-integrity monitoring. Charge amplification, 10 kHz acquisition, and a real-time 512-point FFT pipeline helped separate impacts from fatigue-crack signatures. I took the systems through prototype bring-up, bench characterization, and failure analysis.
I also wrote the sensing architecture, signal-processing methodology, and validation sections of two funded proposals within a seven-person research team.
CSIR-National Institute for Interdisciplinary Science and Technology
Research Associate | Jun 2024 - Mar 2025
At CSIR-NIIST, I fabricated solar cells and tried to understand why one batch behaved differently from the next. I worked on dye-sensitized and perovskite photovoltaics in the indoor-light energy-harvesting program, with the group whose dye-sensitized-cell work later reached a 40% indoor-light conversion-efficiency record.
I built complete devices using FTO substrates, TiO2 photoanodes, thermal processing, ruthenium sensitizers, electrolytes, counter electrodes, and sealed assemblies. Then came J-V curves, IPCE, photoinduced absorption spectroscopy, and XRD with Scherrer crystallite-size analysis. I used controlled experiments to connect processing choices, crystallite structure, spectral response, and charge recombination to efficiency and failure.
Ingenium Naturae
Research Engineer | May 2023 - Jun 2024
At Ingenium Naturae, I built custom plasma and automation equipment for agri-food clients. A specification could start out incomplete; the job still ended with a machine that had to work. I handled system architecture, component selection, fabrication, wiring, firmware, acceptance testing, documentation, and handoff.
The builds involved 20 kV discharge drivers, MOSFET switching stages, PID loops, STM32 firmware, electrode and reactor geometry, Venturi injectors, and production CAD assemblies. I spent plenty of time debugging electrical, thermal, fluidic, and mechanical failures together on the shop floor. Those machines were where a lot of my engineering became real.