👋 Hi, I'm Kevin

I build software
that compounds.

Founding engineer at enttor.ai, building AI outbound at scale · browser automation, OpenAI pipelines, Instagram/LinkedIn prospecting. Previously founding engineer at Samsam. M.Sc. in deep learning at Uniandes.

See work → Get in touch
Available · Q2 📍 Bogotá, CO UTC-5
gen 0 alive 0 fps 0
§01 · About

A small studio of one, mostly happy.

Founding engineer, full-stack, deep-learning grad. Short version below; long version over coffee.

Kevin Gámez en frente de Royce Hall, UCLA K. GÁMEZ · 2026

I'm a software engineer from Bogotá. As founding engineer at Enttor I built the AI outbound engine and browser automation flows for Instagram/LinkedIn prospecting, the Next.js dashboards, the NestJS APIs, and the Supabase + Inngest plumbing that holds it together.

Before that I was founding engineer at Samsam, an e-commerce platform · shopper and merchant apps in TypeScript, React Native, Next.js, Prisma and PostgreSQL. Earlier work crossed data analytics (SQL models, Power BI for HR / risk / operations) and deep learning on satellite imagery for cropland and environmental risk detection.

I finished a B.Sc. in Systems and Computing and an M.Sc. in Information Engineering at Universidad de los Andes, with a specialization in deep learning and a minor in Management. The Game of Life on the right has been alive since 1970; I just gave it a frame.

Based
Bogotá, Colombia
Role
Founding engineer @ enttor.ai
Stack of choice
TypeScript · Next.js · Postgres
Currently obsessed
Browser automation & LLM pipelines
§02 · Tools

The kit, and what I actually reach for.

Everything I've used in production. Coral chips are what I'd pick today, given the choice.

Languages
TypeScript Python Swift JavaScript Java Dart SQL Bash
Frontend
React Next.js React Native Astro Tailwind SwiftUI Flutter
Backend & infra
PostgreSQL Prisma NestJS Node.js Supabase Inngest FastAPI Docker Vercel AWS · GCP
AI / ML
OpenAI Browser automation LLM pipelines Anthropic SDK PyTorch OpenCV YOLO
Data & analytics
SQL modeling Power BI ETL pipelines
§03 · Trajectory

A short, mostly linear story.

Two founding-engineer tours, a master's in deep learning, and the long stretch at Uniandes that taught me how to ship.

Jun 2025 · now

Founding engineer @ Enttor

New York · remote · AI outbound

AI outbound engine · browser automation and OpenAI pipelines for Instagram/LinkedIn prospecting, filtering and automated DMs at scale. Full-stack platform: Next.js dashboards, NestJS APIs, Vercel infra. IG/X/Twitter prospecting on Supabase + Inngest with duplicate-detection logic.

Feb 2024 · Mar 2025

Founding engineer @ Samsam

Bogotá · e-commerce platform

Built shopper and merchant apps in TypeScript, React Native, Next.js, Prisma and PostgreSQL. Refactored core services for ~10% lower response times and shipped an alert routing system that pushed critical-error visibility +70%.

Jan 2024 · May 2025

M.Sc. Information Engineering @ Universidad de los Andes

Specialization in deep learning · concurrent graduate TA

Deep learning, computer vision, applied ML · including satellite imagery for cropland and environmental risk detection. Worked as a graduate teaching assistant designing labs and course materials in parallel.

Jan 2019 · Dec 2023

B.Sc. Systems and Computing @ Universidad de los Andes

Minor in Management · Andrés Bello National Distinction

Five years across systems, algorithms, ML and a management minor. Side projects in Python, TypeScript, Java and Swift; certifications in AWS Cloud Foundations and Cloud Developing along the way.

§04 · Selected work

Things I've built and still believe in.

A handful out of many. Each one taught me a different lesson; the lessons compound, the projects compound.

01 / FEATUREDTypeScript · Postgres · Anthropic · CF Workers

enttor.ai · creative for marketers, at scale.

Pipelines that take a brand brief, generate hundreds of ad creative variants, score them with LLM-based aesthetic + brand-fit evals, and deploy the survivors. From founding line to thousands of MAU; the eval pipeline alone runs ~120k judgments a day.

Read the case study →
enttor
02Rust · WASM

automata-rs

A WebAssembly cellular-automaton sandbox: Conway, Wireworld, Brian's Brain, Lenia. Real-time on a million cells, all running in the browser. The hero on this page uses a sibling implementation.

github →
03TypeScript · GLSL

fractal-atlas

An interactive atlas of escape-time fractals · Mandelbrot, Burning Ship, Julia, Newton · with deep zoom, color schemes, and shareable URLs. A weekend project that ate three weekends.

live →
04Python · numpy

primes-notebook

Number-theory walkthroughs as live Jupyter notebooks: Ulam spiral, Goldbach pairs, the Riemann staircase. Used by a few math teachers I admire.

github →
05Astro · Preact

this site

The page you're on. Open-source, built incrementally, and a place to put math that I think is beautiful next to software I'm proud of.

github →
§05 · Open source

Code I leave out in the open.

The numbers, the languages, and the repos. The contribution graph is decorative; everything else is real.

github.com/kevingamez · live snapshot
0
public repos
school, side projects, ships at work
0
languages shipped
TypeScript leads · Python is second
0
years on github
joined April 2019
language mix across public repos
TypeScript · 33% Python · 27% Swift · 7% JavaScript · 7% Dart · 7% Dockerfile · 7% CSS · 7% Jupyter · 5%

Contributions · @kevingamez

last 12 months
Less More
0
Commits · 12mo
0
Current streak
0
Longest streak
kevingamez / personal-site
This page, source included.
TypeScript
kevingamez / AD_ASTRA2023-SpaceInvaders
Aerial object detection for Amazon deforestation events.
Python
kevingamez / Palladium_Chat
Chat experiment in TypeScript.
TypeScript
kevingamez / budget-app
Personal-finance app, written in Swift.
Swift
kevingamez / GCP-CloudRun
Containerized service deploys on Cloud Run.
Dockerfile
§06 · Now

What's on my desk this week.

A live snapshot. Updated when something changes; usually every Monday morning.

Building
Eval-as-a-service
An API surface so other teams can plug enttor's brand-fit evals into their own pipelines.
Reading
A Mathematician's Apology
Hardy. Slim, beautiful, slightly heartbreaking. Re-read every December.
Listening
Caribou
"Honey" on loop. Good for shipping at 2am.
Thinking about
Knot invariants
Specifically the Jones polynomial. No applications yet, but it keeps me up.
§07 · Bookshelf

Four books that changed how I work.

Not the canon. Just the ones I keep recommending.

Gödel, Escher, Bach
Hofstadter
The Pragmatic Programmer
Hunt & Thomas
A Mathematician's Apology
Hardy
Designing Data-Intensive Applications
Kleppmann

Want to build something compounding?

I take on a small number of consulting projects per year, plus the occasional collaboration. If something here resonates, the inbox is open.

kevingamez.kg@gmail.com → Open dev mode
github / kevingamez x / kevin_gamez linkedin / kevingamez