My foundations are in mathematics, DevOps, and computer science, paired with UX research, writing, and project management.
Robotics projects were my entry into STEM. I started with building an Arduino-based atmospheric sensor to test which chemicals were present in the air at the beach. I also spent more than two years volunteering at the Lawrence Hall of Science, facilitating science and engineering programs and mentoring young visitors.
I am pursuing a B.S. in Information Science and Data Analytics at San José State University, and I earned an A.S. in Mathematics from Berkeley City College. Applied mathematics taught me the problem solving and critical thinking that did much to draw me into computer science.
At InfluxData, I worked on the Product team with the VP of Product (more in Experience). My DevOps experience spans CI/CD automation, Kubernetes deployments, and security pipelines.
My ultimate goal is technical product management, particularly in security or infrastructure. I am pursuing it from two directions, deliberately: product roles that build my management skills, and technical roles in DevOps and security, since not only do I enjoy technical tasks, but I also think having a strong technical foundation is essential for a TPM.
Try it
Tap a node to change its color. Connected nodes can't match, and you only get three colors.
This is my most recent role. You can find my full work history in my resume.
I documented InfluxDB 3, a time series database used by developers on the open-source Core edition, hobbyists on a free at-home tier, and enterprise teams running production systems. I reported to the VP of Product and worked with Product and Engineering to prioritize documentation aimed at developer adoption and faster time-to-value. Below is the project I'm proudest of, followed by more of what I worked on.
The problem: Plugin information came from DevRel, product managers, and engineers in different formats. It was written in one repository but published from another, so every update meant chasing down details and copying them over by hand.
What I did: Connected the two repositories so plugin docs sync automatically, using GitHub Actions, JavaScript, and Python. Helped define a standard README structure and a pre-sync checklist so every contributor knows exactly what to provide. Tested the sync end to end and fixed a validation bug that blocked syncs over warnings alone, so warnings are now reported and only real errors stop a sync.
The result: Documentation updates went from 30+ minutes to 4, with no manual copying errors.
See the merged fix on GitHub →I made 650+ commits to InfluxData's documentation in 2025. Some highlights:
Hands-on engineering work, mostly where something could break, leak, or be exploited if it wasn't built carefully.
Three pieces that show the same instinct from different angles: understand the user, then build or evaluate something that serves them.
I care about writing that a reader can act on immediately, specs, documentation, and reports that don't require a follow-up meeting to decode. Two pieces published on InfluxData's engineering blog:
Breaks down how InfluxDB 3 Core and Enterprise work under the hood: the single-node architecture, flexible "diskless" storage options, the built-in Processing Engine for real-time data transformation, and the read replication and multi-node clustering that Enterprise adds for teams running at scale.
Read on InfluxData →A hands-on walkthrough for standing up InfluxDB 3 Enterprise locally with Docker, pre-configuring the Explorer UI so a team can connect without manual setup, then ingesting and querying sample data, closing with what to consider when moving from a local demo toward production.
Read on InfluxData →An original paper working through graph coloring theory from first principles: proving the greedy algorithm's Δ(G)+1 upper bound, showing how the Welsh-Powell heuristic tightens that bound in practice, and walking through Brooks' Theorem using Lovász's vertex-ordering argument. Ties the math to real systems, compiler register allocation, CI/CD job scheduling, and procedural generation all reduce to the same coloring problem.
Read on GitHub →A plain-language breakdown of a published neuroscience paper on geometric hallucinations, explaining how mathematical models of the visual cortex, built on tools like Euclidean symmetry and perturbation theory, can account for the geometric patterns people see during migraines or under certain psychoactive substances, even in total darkness.
Read on GitHub →Leadership and coordination work outside of paid roles, most of it was really just project management before I had the title for it.
A quick snapshot of the tools behind the work above. If you want the full picture, roles, dates, and results, you can download my resume.
I'm always happy to talk about roles, projects, or how any of this was built. LinkedIn is the best way to reach me.