Software quality · AI · Mobile apps

Lorenzo Perrone

Software QA and IVVQ Engineer, AI Champion, and mobile app developer with 16+ years of experience in biometrics and IT. I solve complex quality problems and build useful, privacy-conscious products.

LinkedIn Email GitHub

Base: Turin, Italy · Working with international teams
Focus: Software quality, support leadership, biometrics, and privacy-aware mobile products
Approach: Clear systems, dependable releases, honest communication, and practical product decisions

About

Software quality, reliable infrastructure, and technical support built around people.

For more than 16 years, I've worked where product reliability meets real user needs. I test biometric SDKs and multimodal devices, investigate difficult field issues, support international teams, and manage IT infrastructure. That experience also shapes the mobile apps I build: focused tools with clear interfaces, no unnecessary tracking, and no dark patterns.

Experience

16+ years in software quality, biometrics, technical support, and IT.

Multinational Tech Corporation

2011 - Present

AI Champion

2026 - Present

I support AI adoption for around 50 colleagues in R&D through practical training, clear guidance on company policies, and hands-on use of approved internal tools.

Software QA & IVVQ Engineer

2014 - Present

I test biometric SDKs and multimodal scanners, and develop Python and C++ tools for automated diagnostics and log analysis. My work includes anti-spoofing certifications such as iBeta PAD and border-control validations for the EU Entry/Exit System.

Head of Support

2018 - 2024

I led global third-level technical support and root-cause analysis for complex field issues. I also managed onboarding for major government customers and led the migration to ServiceNow after a corporate acquisition.

IT Manager

2016 - 2024

I managed VMware, Cisco, and SonicWall infrastructure and coordinated the post-acquisition integration of IT networks. I also reviewed GDPR compliance and data-retention policies for internal biometric datasets.

Selected apps

Independent mobile apps designed to be useful, calm, and private.

From accessible color recognition with Argyo and private emotional journaling with HicNow to offline Italian card games, each app solves a specific problem with clear interfaces, transparent behavior, and no advertising.

Local-first emotional grounding

HicNow

A private emotional journal to pause, name the moment, and keep it close.

HicNow maps your current emotional state on a continuous model inspired by Russell, then turns it into a private capsule with text, audio, photos, or video. The calm capture flow supports emotional grounding, while every memory stays stored locally on your device.

Grounding map — A continuous mood selection flow designed to reduce friction and invite calm reflection.
Private media capsules — Each moment can be stored with text, audio, photo, or video and kept entirely on-device.
Local-first archive — Capsules, preferences, reminders, and media are designed to remain on the device by default, without requiring an account or a cloud backend.

Current status Available on iOS. Coming soon on Android.

App Store

HicNow Privacy

Accessible color analysis

Argyo

Identify and describe colors with confidence, from your camera or photo library.

Argyo is an accessible color recognition app for camera input and gallery images. As a colorblind developer, I built the tool I could not find: clear color names, HEX values, and a contrast-aware interface that never relies on color alone. No ads and no tracking.

Gallery and camera flows — The MVP covers both static image analysis and live camera sampling with accessible overlays.
Evocative naming — Argyo prefers memorable color names before technical labels to make daily communication easier.
Accessibility-first rules — Every state combines color, icon, and label so the interface remains understandable for every user.

Current status Available on iOS. Coming soon on Android.

App Store

Argyo Privacy

Classic Italian Scopa, no noise

Pure Scopa

Play classic Italian Scopa and 4-player team Scopa offline, against AI opponents.

Pure Scopa brings classic Italian Scopa and its 4-player team variant to mobile. The AI plays offline and receives cards under the same rules as the player. There are no accounts, ads, tokens, or energy systems, and matches, settings, and statistics remain on your device.

Classic Scopa and 4-player team Scopa — Two classic modes with traditional scoring: cards, denari, settebello, primiera, and scope.
Offline AI with three levels — Three difficulty levels — easy, normal, expert — that change how the AI plays, never the cards it receives.
Local-first by design — Save/resume, settings, and statistics live on the device. No account, no cloud backend, no telemetry.

Current status Available on iOS. Coming soon on Android.

App Store

Pure Scopa Privacy

Classic Italian Briscola, no noise

Pure Briscola

Play classic Italian Briscola offline, with five AI difficulty levels.

Pure Briscola brings the classic Italian card game to mobile with five levels of offline AI. The cards are dealt under the same rules at every difficulty. There are no accounts, ads, tokens, or energy systems, and matches, settings, and statistics remain on your device.

Classic Italian Briscola — Authentic two-player Briscola with traditional scoring and trump mechanics.
Offline AI with five levels — Five difficulty levels — from beginner to expert — each with distinct strategies that improve play without cheating.
Local-first by design — Save/resume, settings, and statistics live on the device. No account, no cloud backend, no telemetry.

Current status Available on iOS. Coming soon on Android.

App Store

Pure Briscola Privacy

Labs & Open Source

Open-source projects and experiments.

Here I collect tools, prototypes, and ideas that I develop over time and share openly.

Open Source / Experimental AI

Subvocal

An experimental tensor-native agent for local AI.

Subvocal is a Node.js experiment that bypasses conventional HTTP and JSON layers to communicate directly with llama.cpp. It explores shared-memory dual-agent architectures and direct KV-cache manipulation for local LLM workflows. It is a research prototype, not a production tool.

View on GitHub

Open Source / Developer Tools

Promptino

Your tiny teleprompter overlay.

A lightweight, unobtrusive teleprompter that stays on top while you work or record. Paste your script, adjust the speed, and focus on what you want to say.

View on GitHub