Fractional CTO · Traditional industry · AI engineering
Most technical leaders know software or industry operations — rarely both. I’ve built the infrastructure behind a 100-developer fintech, cut Amazon Device OS’ release lead time from days to hours, designed ML infrastructure at LandingAI, and I put that same stack to work inside manufacturing and other traditional businesses.
I join a few companies at a time, hands-on, 10–20 hours a week. You get someone who has shipped distributed systems and AI — and who understands how traditional industries actually run.
No prep, no pitch. You’ll leave with a clearer picture of your options either way.
Taking one AI diagnostic at a time
See the problems I help with Fixed-price AI diagnostic
Y Combinator
01 / How I can help
Operations, engineering, and technology leaders at traditional-industry companies — manufacturing, distribution, logistics, industrial — usually come to me in one of three situations. Here’s what I do about each.
01 / You have engineers, but no one steering.
You shouldn’t have to referee architecture debates or guess whether to build or buy. I connect the technical roadmap to your business priorities, make those calls with you, help you hire the right people, and coach your engineers and managers to take ownership.
Technical strategy · Key hires · Coaching
02 / Releases are slow and things keep breaking.
Slow releases, fragile infrastructure, and manual work are eating your team’s time. I work in the systems alongside your engineers to improve delivery, reliability, and security — including the data and ML infrastructure behind the tools your operation depends on.
Architecture · Deployment automation · Data & ML systems
03 / You need a working tool in the operation, not another slide deck.
Define the smallest useful system and build it together — an internal tool, a data integration, a workflow that removes manual work. I stay close to the code and focused on what the business needs to run better next.
Scope · Hands-on development · Integrations
02 / Experience in practice
The situations above aren’t theoretical. Here’s what solving them has looked like.
Kueski / Infrastructure & engineering leadership
100+ developers
Built the infrastructure and DevOps function from a one-person operation into a multi-team organization. Established automated, security-focused infrastructure for a regulated fintech serving more than 100,000 users.
Amazon / Device OS
Days to hours
Led a project to reduce lead time for changes in the Device OS organization, helping engineers move work through the development lifecycle faster.
Keep Technologies / Banking
~40%
Led the project to create a new payment channel that generated millions in revenue and went on to receive roughly 40% of customer payments. Helped restructure and relaunch teams, then led Risk and Growth work focused on application hardening and new growth opportunities.
AcerosVS / Consulting / Engineering director
>20,000
Built and now lead the software infrastructure and internal tools supporting a steel business selling more than 20,000 tons per month. From consulting partner to engineering director, I’ve helped build the systems behind that growth and am now integrating AI into the company’s technology strategy.
03 / Who you’d be working with
I’m at my best where software, infrastructure, and people meet. I can work through an architecture decision, get into the code, and help the team develop the judgment to own it.
As co-founder and CTO of MovingLake, a Y Combinator company, I built real-time data integration technology with Rust, Go, and Kafka. At Andrew Ng’s LandingAI, I helped design and implement infrastructure for large-scale machine learning inference. Those experiences inform how I approach data systems and AI today.
My foundation is electrical engineering at Tecnológico de Monterrey, Campus Ciudad de México, where I graduated with the highest GPA in my program. My career since has taken me from early-stage companies to Amazon, with consulting work alongside operating roles.
Tools I work withRust · Go · Kafka · AWS · React
Explore my open-source projects04 / Working together
You work directly with me, remotely, for 10–20 hours a week. We agree on priorities and scope: ongoing fractional CTO support, a focused platform project, or a product build.
We talk about your customers, goals, and what’s getting in the way. No prep needed. If I’m not the right fit, I’ll say so and point you somewhere useful.
We agree on priorities, deliverables, and what success looks like. You’ll understand the tradeoffs and what we’re deliberately leaving for later.
I work in visible increments and share the reasoning behind decisions. Your team gets the code, documentation, and context to keep moving after the engagement.
05 / Let’s talk
“Very few people can move as fast as Edgar does without cutting corners. As our CTO he was the technical backbone of MovingLake — from designing our real-time data pipelines to writing low-latency integrations in Rust and Go, he built the core of what we shipped. Any founder would be lucky to have him in their corner.”
Tell me what you’re building and where it’s stuck. In 30 minutes you’ll have an outside read on it, whether or not we work together.
Taking one AI diagnostic at a time
Not ready for a call? Email me the problem in three lines.
hello@edgarcabrera.comRunning a traditional-industry company and wondering where AI belongs?See the fixed-price AI diagnostic