open_to_opportunities | remote · argentina

Language
is the Interface

Localization QA Specialist. Ten years building linguistic, cultural & technical systems that ensure Spanish Latin American content feels native, from AAA games to global fintech. I also build the tooling that keeps QA honest at scale.

→ get_in_touch see_work.md

~/qa-engine · node

● live

$qa run --lang es-419 --source ./strings

› loading translation memory… ok

› parsing 2,418 strings… ok

› evaluating 6 dimensions… ok

{

"linguistic_accuracy": 98.7,

"cultural_score": 99.2,

"terminology": 97.4,

"register": "native",

"error_detection": +35%,

"verdict": "ship"

}

yrs_exp

words_reviewed

0M+

error_detect_gain

0%↑

§ 01 / capabilities

Where language meets precision.

I don't just review text. I build systems (tooling, rubrics, pipelines) that ensure every localized word performs at the highest standard. Chrome extensions, GPT evaluators, MQM-style reports, all humming in the background while the language does its job.

//qa_intelligence.ts

LIVE

linguistic_accuracy

98.7%

cultural_score

99.2%

error_detection

+35%

avg_turnaround

4.2h

> review_stream · active

14:02onboarding/register.es-419 validatedpass

13:48tone mismatch · tutorial.str[041]fix

13:31terminology aligned · TM v2.1pass

13:19gender agreement · checkout.btn[03]fix

//review_scheduler.cron

AUTO · WEEKLY

weekly QA cycle · active days highlighted

sunoff

monreview

tuereview

wedbuild

thureview

frisign

satoff

automated cycle across active client projects. review days highlighted · save_protocol().

§ 02 / builds

Tools I've built to make QA scale.

A linguist who ships software. Each of these started as a real bottleneck in localization QA and became a working product: every one runs on privacy-first architecture, because client material under NDA should never leave the machine.

[01] browser · desktop

QA Checker

Linguistic QA for bilingual files that runs entirely on your own machine: no uploads, no server, works offline.

17 automated checks across completeness, consistency & content integrity
Locale-aware number comparison & source-aware punctuation: kills the usual false positives
Reads XLIFF, SDLXLIFF, MQXLIFF, TMX, TBX, PO, TSV & CSV; exports HTML / Excel / XML
Imports existing industry checklist files, so accumulated client rules carry over
Glossary-driven key-term checks + checklists with regex, wildcard & boolean search
Same engine ships as a native desktop app

TypeScript · React · Tailwind · Vite · Tauri (Rust)

[02] web app

DTP QA Reviewer

Automates post-DTP bilingual PDF review, catching translation, layout & visual-callout errors before they reach the client.

Compares source & translated PDFs, streams a structured QA report in minutes instead of hours
Three-pass method: structural scan → page-by-page issue table → visual-spatial callout mapping
Flagship check verifies each diagram label's leader line points to the right physical element
~275-line specialist review prompt with per-locale rules (es-LA / pt-BR)

React · TypeScript · Node/Express · frontier LLM API (streaming) · Docker

[03] web app

LinguaAudit Pro

AI-powered LQA platform for EN → ES-LA, built on MQM / ISO 17100 standards.

Multi-model consensus: fans each audit out to three independent LLMs, then dedups & attributes every finding
MQM score (0–100), segment-linked error table, inline issue → fix diffs
Deterministic prompting (temp 0): the same audit returns the same errors twice
Client-side XLIFF 1.2 / 2.0 parsing anchors each error to its segment ID

React · TypeScript · Node/Express · multi-provider LLM orchestration

[04] macos · offline

Traductor EN → ES-LA

A local translation workbench that runs fully offline: zero data ever leaves the machine.

Native macOS app powered by a local open-weight model: no cloud, NDA-safe
Three modes (transcreation, UI precision & legal), each with its own prompt, glossary & temperature
Global shortcut translates selected text from any app in one gesture
Anti-hallucination guardrails: few-shot anchoring, hard length caps, low temperature

Python · FastAPI · local LLM runtime · PyWebView · PyObjC

[05] chrome extension

Super Boost!

A productivity suite injected into an enterprise translation platform that ships with no plugin system or API.

~10,000 lines of JS/CSS: works entirely through runtime DOM injection & MutationObserver
13+ automated QA checks with inline error badges & team-maintained custom rules
Four inline search/research panels + AI-powered cross-segment consistency review
Built solo for a team of es-419 and pt-BR linguists at a major technology company

Vanilla JS · Chrome Extensions API (MV3) · spreadsheet API · LLM API · ICU MessageFormat

§ 03 / philosophy

Two questions. One bar.

the industry asks

"Is it
translated?"

A checkbox answer. Volume, speed, delivery. Often enough to ship. Rarely enough to resonate.

I ask

"Does it feel
native?"

A craft answer. Rhythm, register, the pulse of how people actually speak. That's the bar I hold every string to.

§ 04 / experience

Built through a decade of deep work.

[01]

Senior Localization Specialistfreelance · remote

End-to-end linguistic, functional & cultural validation. Tooling, reports, review cycles.

2015 → now

↗

End-to-end localization QA for web, SaaS & product content
Sanity & regression testing across Web / iOS / Android
Built Chrome extensions integrating spreadsheet-based QA rules into CAT tools
Developed GPT-powered QA systems evaluating six linguistic dimensions
Delivered MQM-style reports with actionable feedback

[02]

Localization QA & Solutions LeadRoundTable Studio

Quality strategy, AI-assisted pipelines & proprietary QA tooling across accounts.

2015 → now

↗

Quality Strategy: design measurable quality frameworks & KPIs to track trends, reduce recurring errors, and drive continuous improvement across accounts
AI & Automation: architect and maintain AI-assisted feedback pipelines, automating file comparisons and validation to scale efficiency
Tool Development: build and document proprietary QA solutions, custom extensions & scripts to eliminate single points of failure
Account Leadership: primary quality point of contact, overseeing translation/review workflows and data-backed communication for client QBRs
Mentorship & Knowledge: lead knowledge transfer, team training & best-practice adoption to grow the team and quality awareness

[03]

AI Language Specialistmultiple vendors

Evaluated AI/MT output. Prompt evaluation, LLM benchmarking, AIPE pipelines.

2020 → now

↗

Evaluated AI-generated content and MT output for fluency, accuracy, cultural fit
Participated in AI training, prompt evaluation and structured QA workflows
Supported LLM benchmarking and AIPE quality assessment pipelines
Ensured culturally inclusive Spanish dialogue across UX and moderation projects

§ 05 / toolkit

The tools behind the craft.

A technical linguist with language expertise & engineering capabilities to build smarter QA workflows.

localization & qa [01]

LQA
linguistic QA
web & product loc
UI/UX loc
SaaS loc
translation
proofreading
cross-platform testing
bug tracking
TM management
terminology

tms / cat tools [02]

Smartling
Phrase
XTM
Trados
MemoQ

pm [03]

JIRA
Asana
Workfront

automation & dev [04]

JavaScript
Chrome Extensions
Cloud APIs
HTML/CSS
REST APIs

ai & machine translation [05]

AI-based QA
prompt engineering
LLM eval
MT eval
MTPE
TEP
AIPE
QA benchmarking

languages [06]

es · native
en · C2
pt · prof

§ 06 / contact

Make your product
feel_native().

Whether you need a Localization QA Specialist or someone to build automation to scale your QA processes, let's connect.

→ ezequiel.moreno5@gmail.com linkedin ↗

Language is the Interface