Vince Linero · Software Engineer
Every role a product needs.One engineer.
attention, undivided.
I’ve covered the whole product lifecycle for 15+ years — business strategy, architecture, UX, front end, back end, deployment — since long before AI could write a line of code. AI didn’t give me that range; it multiplied it.
Today that means taking a product from idea to production — increasingly LLM-native products, where the model does real work inside the system. And when it’s time to scale, I can integrate and train the technical team — on an AI-assisted dev workflow.
engineer · seeker
I aim for perfection — but recognize when good enough is good enough.
The whole lifecycle
Getting a product out the door usually takes a product manager, a designer, developers, and someone on ops. I can cover the full sequence myself — and, with the help of AI, deliver much faster.
- 01
Understand
The business need behind the feature list — sitting with the relevant stakeholders to define requirements: what the product must prove, and to whom.
- 02
Plan
Scope, architecture, and the shortest path to working software.
- 03
Design
UX and UI shaped around how people will actually use it.
- 04
Build
Front end, back end, and everything between.
- 05
Deploy
Infrastructure, CI, and the ops to keep it running.
- 06
Harden
Testing, documentation, iteration — until the product is ready for real users and the codebase is ready to grow.
- 07
Scale
When the product calls for a team, I integrate and train one — adjusting the AI-assisted dev workflow that works for one engineer into one that works for many.
- 08
Support
I stay with the product once real users are in the system — iterating, supporting, leading the workflow.
I work autonomously — once I commit to something, I deliver. I report continuously as I go: plain-language progress updates, decisions surfaced early. And I stay within reach — as present for the quick question as for the deep work. Trust follows from there.
Case studies
The work, told plainly
Three engagements, most recent first. These products live inside private client systems — no screenshots to show — so each story stands on what was asked, what shipped, and what it took.
Case study · 2025 – present
Rebuild Better™ — one engineer, a whole platform
EcoClaim helps the property-insurance industry measure and reduce the carbon footprint of restoration work. I designed and built its platform — from the first conversation to production — as the sole engineer.
The brief
Turn the estimates the industry already produces into environmental intelligence — without an engineering team to build it.
What shipped
An LLM-native pipeline — structured extraction, prompt pipelines, evals — that reads restoration estimates, recognizes what will be rebuilt and with which materials, and quantifies the carbon impact, surfacing lower-carbon alternatives with cost and emissions savings side by side.
The discipline
Insurance runs on trust, so every model decision runs inside governed, auditable workflows — evaluated against known-good data, with deterministic logic keeping the AI honest.
The scope
UX/UI, front end, back end, background processing with real-time front-end updates, multi-tenant access, dashboards, reports and exports, an MCP server exposing platform data to AI assistants — plus deployment, CI, and ongoing operations.
The stack
Vue 3 on the front end; Symfony with GraphQL and event sourcing on the back; a Python FastAPI service running the LLM pipeline; MySQL and Redis underneath.
The result
In production, serving restoration contractors and insurers across multiple regions.
Every stage of the lifecycle, exercised on one real product — with discipline carried from four years inside a regulated exchange.
Case study · 2020 – 2024
ACX — from first commit to VP of Engineering
ACX (AirCarbon Exchange) is a digital exchange for environmental commodities — carbon credits traded like financial instruments. I spent four and a half years there, Singapore then Abu Dhabi. It’s the chapter that reshaped my career.
The setting
A global exchange serving corporates, financial traders, and carbon project developers — spot trading, auctions, and settlement, running on distributed-ledger rails.
The start
Two engineers. I built the Exchange frontend v1 single-handedly — the trading screens clients actually traded on — in ReactJS against a Node.js backend.
The scale
The team grew to eight engineers, and I grew with it — the last three years as VP of Engineering. I kept shipping code on every front the whole way; staying on the tools was how the team moved faster.
The stakes
In Abu Dhabi the exchange became regulated — licensed by the FSRA under ADGM. That means financial-grade infrastructure: real money, real compliance, real audits.
The stack
A ReactJS trading front end; Node.js/TypeScript microservices; a C#/.NET matching engine; FIX connectivity for institutional traders — including regulatory drop copy to the FSRA; distributed-ledger settlement, on AWS.
The result
Environmental Finance named ACX Best Carbon Exchange four consecutive years — 2021 through 2024 — the span in which this platform was being built.
Where building the product and building the team stopped being different jobs — though holding a whole business in one head was a skill from an earlier chapter.
Case study · 2013 – 2018
Jetsemani — the whole business, not just the code
A hospitality marketplace in Colombia, founded by Tom Herman — one of the lead characters of the documentary Startup.com. He recruited me as a developer. I ended up running the company.
The inheritance
A PHP platform a couple of years old, built by an outside agency, with hospitality inventory but no real traction.
The rebuild
React was brand new; Airbnb was betting on it, and I did too. I learned it and rebuilt Jetsemani.com from scratch. I’ve been early to the promising bet ever since.
The growth
I became General Manager while still building the product: sales, vendor relationships, and a multidisciplinary team of eight — developers, design, sales, business development.
The reality
The market was crowded, and the marketplace didn’t win. We pivoted to building sites for the hospitality industry; after nearly five years I was spent, and closed my chapter.
What it left me
The ability to hold an entire business in one head — and a hard lesson in what attention costs when it’s divided. That lesson became a practice.
Everything above was built with what this chapter taught.
Personal project · in daily use
Maia
Multi-Agentic Intelligent Assistant
The one thing on this page I can actually show you.
Taking the form of an AI software factory. Built end to end, for my own use. Not a product. The workshop.
Requirements met or not, evidence attached, and a decision waiting for a human. Nothing here is asserted — it is derived.
Requirements → specs → plan → build
- Requirements met or unmet, decisions logged — with the specs, plans and work behind each.
- Plans that execute themselves — tasks, worktrees, a validator that gates the merge.
Knowledge base · memory
- A knowledge base and memory per project — built as it goes.
QA
- A built-in browser and CLI tools — end-to-end QA, by the system itself.
Notifications
- A notification only when a decision is stuck waiting on me.
Any agent CLI
- Agent-agnostic — a web application that renders real terminals, so Claude Code, Codex or any agent CLI can run in it.
Next.js · TypeScript · Python / FastAPI · MongoDB · ChromaDB · WebSockets · tmux · Claude Code, Codex or any agent CLI as runtimes · an internal model gateway across Anthropic, OpenAI and OpenRouter · local inference on an NVIDIA DGX Spark via Ollama · self-hosted.
What I bring
The full range of what a product needs — and the ways of working that carry it.
Technical
Authentication & SSO
OAuth · JWT · roles & permissions
Real-time & background work
WebSockets · Redis · Arq
Management screens & admin tools
dashboards · search & filters · bulk actions
Reports & exports
CSV · Excel · PDF
Data modeling & storage
SQL · MongoDB
AI agents & automation flows
model-agnostic orchestration · MCP · evals
AI governance & auditing
traceable decisions · logged actions · human review
Documented for AI, not just humans
platform, workflow & setup docs — vibe-code on top
Interfaces people enjoy using
UI/UX — curating what AI designs
DevOps
CI/CD · Docker · cloud & self-hosted
Personal
New domains, learned fast
hospitality · carbon markets · insurance · property
Plain-language communication
stakeholders · users · teams
Teams, integrated and trained
AI-assisted workflow · reviews · docs
Presence under pressure
incidents · deadlines · pivots
Meditation & yoga, practiced
a standing discipline of attention
Perception before reaction
what’s said · what’s meant · what’s missing
Patterns, noticed early
in systems · data · people
Thinking that goes to the root
first principles · root causes
“He can think strategically, design UX/UI for a product, code the front end, code the back end, and supply dev ops. If you need a really full stack engineer who can operate independently, Vicente’s terrific.”
“Vicente’s ability to translate complex business concepts into seamless software workflows is awe-inspiring. He is a skilled engineer and a strategic thinker.”
the practice
I sit in silence every morning before I touch a keyboard.
Years of meditation and yoga are not a footnote to the engineering — they are where the patience for hard problems and the taste for simplicity come from. The geometry at the top of this page breathes at the pace of a resting breath. The work is made with the same attention.