About

I'm Marisniulkis Lescaille — Mari. I'm a distributed systems engineer, and I write about what I learn along the way.

I grew up in Cuba. My first business, in high school, was collecting glass bottles door-to-door during the economic crisis and trading them in for cents apiece. My second was riskier: after college, two friends and I started a small software consulting firm at a time when that was illegal in Cuba. We built a statistical information system for one of the country's biggest biomedicine institutions and shipped content-management projects to Chile before the risk outgrew the reward and we each found our own way out.

Before I left, I co-founded VenusIT, a community to make women in computing visible at my university, and with my co-founders and the support of ACM-W, we organized the first Grace Hopper Celebration of Women in Computing ever held in Cuba.

I came to San Francisco in 2017 through the Silicon Valley Internship Programme. Two years later I was on the other side of the table, volunteering as an interviewer for the same program. In between, I spent three years at Survata (now Upwave) doing the unglamorous work I love most: migrating production databases nobody wants to touch, building shared libraries and documentation culture, and designing the code-reading interview the team still used after I left.

While doing all that, I ran VoiceFirst Weekly, a daily media company about voice assistants, back when "Alexa, good morning" still felt like science fiction. Newsletter, podcast, and a daily flash briefing shipped on Alexa, Google Assistant, Spotify and Samsung Bixby, where I was an early developer partner. I interviewed founders on conference floors with a phone and a lavalier mic. It taught me how platform waves work from the inside: I've learned so much by building on the platform and reporting on it at the same time.

Since 2020 I've been at Netflix, where I'm now a staff software engineer.

Through all of it: I take notes on everything, I keep the receipts, and I believe debugging, reading code (yes, even in the age of AI) and evolving software successfully are some of the most underrated skills in software. That's what I write about here. These days I'm especially interested in what the AI era does to that skill: when machines write more of the code, the people who can read systems, question them, and debug them matter more than ever.

Outside of work I write poetry (sometimes with algorithms).

If any of this resonates, say hola.