A career can look like a straight line only after somebody has taken an eraser to it. Eleazar Vega-Gonzalez's public record resists that tidying. It begins with a wearable journal assembled from GPS, Bluetooth, an Android phone, and the earnest hope that software might remember your life for you. It runs through Microsoft patents, an attempt to loosen the clock's grip on computer chips, a checkout button meant to make credit cards optional, and an AI app that turns prompts into branching adventures.
The nouns keep changing. The verb is mostly the same: simplify. Vega-Gonzalez repeatedly moves beneath the visible product, finds an awkward system, and asks how much of its machinery the user can be spared. His LinkedIn biography compresses the taste behind that work into a tiny manifesto: he loves “ai + slick ux” and hates “bad software.” Lowercase, clipped, and almost comically direct, it is also a useful map.
First, a machine for remembering
At Carnegie Mellon in 2012, Vega-Gonzalez was part of a four-person embedded-systems team that built lifeLogger. The name belonged to an era when attaching “Logger” to a gadget still sounded futuristic. The concept was more ambitious than the hardware's homemade silhouette suggested. A small wearable device collected location data, spoke to an Android phone over Bluetooth, and sent context to a web service. The website could show where a person had been and which friends had been nearby.
Vega-Gonzalez implemented the REST API and the site that visualized the data, working with node.js, Amazon Web Services, MySQL, and Google Maps. The team's poster said the prototype could sync quietly in the background after setup. Even then, the interesting part was not the sensor. It was the disappearance of effort. A journal normally asks you to stop living and write. lifeLogger tried to let the system do the clerical work.
There was already a playful streak around the engineering. While at Carnegie Mellon, he and classmate Saxon Parker contributed a keybinding feature to the open-source Minecraft X-Ray project. In 2014, after a friend used a hashtag he did not understand, Vega-Gonzalez spent a Sunday making a Chrome extension called What the Hash? Annoyance met code; the annoyance lost.
The invisible argument inside a chip
After Carnegie Mellon, Vega-Gonzalez worked in Microsoft's orbit. Patent filings name him among inventors on systems for capturing web content, sharing captured assets, and observing data across restricted application containers. One of those patents, Data Object Observation Among Domain-Restricted Containers, was issued in 2018. The subjects are technical, but the theme is familiar: help information pass cleanly through boundaries that are complicated underneath.
In 2015 he co-founded Reduced Energy Microsystems with William Koven and Dylan Hand. REM was a fabless semiconductor company, meaning it designed chips without owning the factories that made them. Its target was low-power computer vision and AI at the edge, in devices such as cameras, drones, wearables, and robots. Y Combinator accepted REM into its Summer 2015 batch and described it as the accelerator's first chip startup.
REM's technical bet concerned the clock that coordinates a conventional processor. Synchronous chips make components wait for the next tick, even when some work finishes earlier. REM pursued asynchronous designs that could let operations proceed at their own pace and change voltage according to demand. In plain English, the chip could sip power for easy tasks and wake up for harder ones.
The proposition was not merely a science project. REM emphasized internal tools that could work with familiar design flows and present a familiar processor to programmers. Vega-Gonzalez told TechCrunch that the tooling was what could let a startup move quickly against established chip companies. The product promised a difference people might feel as longer battery life or greater local intelligence, while leaving the difficult circuitry out of sight.
In 2017, REM announced a $2 million seed round associated with Draper Associates. It received a $224,576 Phase I award from the National Science Foundation for an automated design flow for energy-efficient vision and machine-learning chips. It also joined the RISC-V Foundation. Vega-Gonzalez praised the architecture's openness, a practical choice for a small company that needed room to build and contribute without asking a gatekeeper for permission.
The same instinct, four different layers
Then the card became the problem
Vega-Gonzalez's next company looked unrelated. In 2022 he started Coil with Ryan Hall, a friend he had met at Carnegie Mellon more than a decade earlier. They entered Y Combinator's Winter 2023 batch with a pay-by-bank product for merchants. The pitch took aim at card processing fees and the recurring payments that fail when a card expires, changes, or is declined.
The first product was a “Pay with coil” button that could sit inside an existing checkout. A new customer linked a bank account through Plaid; returning customers could verify themselves through a code. Behind that button were bank transfers, authentication, and payment infrastructure. In front of it was an interaction intended to take seconds. Once again, the heavy engineering stayed backstage.
Coil's launch copy was blunt about its enemy, and Vega-Gonzalez said he and Hall wanted to help businesses take control back from card networks. He also described Y Combinator's impact on him in unusually personal terms. Returning to the accelerator eight years after REM suggested continuity even as the material changed. The first visit involved a chip measured in volts. The second involved money moving between accounts.
A plot twist called Coil
The current Coil supplies its own plot twist. Its public terms now describe web, iOS, and Android applications for interactive storytelling. A person gives the platform a prompt or another input; the system generates narrative, images, and choices that shape what happens next. The company Y Combinator once introduced through payment fees is now listed in AI, gaming, and media.
A pivot this wide can tempt an observer to invent a dramatic conversion. The public facts support a quieter reading. Vega-Gonzalez has long built around AI, efficiency, and polished interaction. REM wanted intelligence to happen locally without exhausting a battery. Payments Coil wanted a bank transfer to feel as easy as a button. Storytelling Coil wants generative models to feel like play. The market changed, but the interface remained the place where the argument was won or lost.
There is some comic evidence that this builder was never going to remain entirely inside an engineering diagram. At Phillips Academy, the teenage Vega-Gonzalez played a gang leader called Big E in a black-and-white student movie about a ridiculous local feud. IMDb later recorded an associate-producer credit on the 2012 short First Kiss. The credits are modest, yet they make today's storytelling turn feel less alien. Technical seriousness and theatrical mischief have shared the same résumé for a while.
The useful thing to steal
Founders are often advised to become legible: one market, one identity, one neat sentence. Vega-Gonzalez's path offers a different kind of legibility. Look below the market labels. Each project began near a piece of friction he considered needless. Manual journaling. Wasted processor power. Card fees and failed recurring charges. The blank page between a person and a story.
His career also shows the value of small prototypes beside large ambitions. lifeLogger was a capstone with off-the-shelf boards and a phone. What the Hash? consumed a Sunday. REM required specialist silicon knowledge, venture capital, and federal research support. Coil's current platform sits on modern generative models. Scale varied wildly. The reflex to make something did not wait for the perfect size of problem.
A useful exercise emerges: keep an irritation ledger. Write down every moment when software asks a person to understand the machine's internal organization. Notice which annoyances return. Some deserve a shortcut, some a weekend extension, and a rare few a company. The list may reveal a founder's taste more honestly than a mission-statement workshop ever could.
Eleazar Vega-Gonzalez has not published a grand unified theory of his choices. He has left a series of working examples. The newest one invites a user to choose what happens next. That is a fitting product for a career whose coherence appears only after several improbable turns. The plot keeps moving. The hand on the interface is recognizably the same.