Building useful software in Cagliari 2026 / Portfolio

I turn hard
problems into
clear products.

Software engineer working across applied AI, native macOS apps, and data-rich interfaces. I care about the whole path—from a defensible idea to software people can actually use.

Explore
the work

01 / Selected work

Proof, not promises.

Six projects selected for technical depth, completeness, and the quality of the problem they solve.

Research engineering2025

MUVAD

Explainable video anomaly detection, rebuilt as one coherent multimodal system.

My thesis project unifies visual understanding, reasoning, anomaly scoring, and natural-language explanations in a single inference pipeline—with an interface for exploring the results.

83.17%AUC on UCF-Crime
−82%memory vs prior pipeline
2Bcompact multimodal model

Read the repository ↗

AI Capability Signals analytical dashboard

Live report ↗

Data engineering / research02

AI Capability Signals

A reproducible evidence system for comparing frontier models without collapsing incompatible benchmarks into a fake universal score.

PythonData pipelinesStatistical audit

View source ↗

macOS app ↗

Native product03

Rapid Reader

A focused macOS reader for EPUB, PDF, documents, and web articles—combining RSVP speed with a complete text reading experience.

SwiftUISwiftPMLocal-first

Download release ↗

macOS utility ↗

Systems / visualization04

DiskVisualizer

A native, cancellable disk scanner with treemap navigation, duplicate candidates, extension analysis, and Finder integration.

SwiftUIConcurrencyCharts

Download release ↗

Maestro di Orchestra agent orchestration dashboard

Developer tool ↗

Developer experience05

Maestro di Orchestra

A local dashboard and CLI for orchestrating Codex and Cursor Agent runs with isolated workspaces, review gates, event logs, and secret redaction.

ReactCLIAgent orchestration

View source ↗

Civic data product06

Ciclismo Sicuro

A comparable dashboard for all 106 Italian provincial capitals, with explicit methodology, missing-data policy, adjustable weights, source provenance, and auditable coverage.

ReactTypeScriptOpen data

↗

Gabriele Monni at the University of Cagliari graduation

Gabriele Monni
Cagliari, Sardinia

02 / About

Curious by nature.
Rigorous by choice.

I’m a software engineer with a degree in Applied Computer Science and Data Analytics from the University of Cagliari.

I like problems that sit between research and product: understanding a model, making the data defensible, then designing the interface that makes the result useful.

My work moves comfortably between Python and TypeScript, Swift and SQL, experiments and production constraints. The common thread is clarity: in the architecture, the evidence, and the experience.

109/110 BSc, Applied Computer Science
& Data Analytics · 2025

03 / Experience

Where I’ve been useful.

APR 2026 — NOW

Current role

IT Technical Officer

University of Cagliari · Full-time · On-site

Cagliari, Italy

2022 — 2025

BSc in Applied Computer Science
and Data Analytics

University of Cagliari · Final grade 109/110

Cagliari, Italy

2021 — 2024

Boat Rental Assistant

Aequor srl · Seasonal

Baunei, Italy

04 / Capabilities

From first question
to working system.

01Applied AI & multimodal systemsModel integration, evaluation, local inference, explainability
02Data products & evidenceReproducible pipelines, source provenance, analysis, visualization
03Product engineeringPython, TypeScript, React, SwiftUI, SQL, testing and release workflows
04Technical communicationClear interfaces, rigorous documentation, Italian and English

05 / Contact

Have a hard problem?

Let’s make it clear. ↗

Based in Cagliari, Italy
Open to ambitious teams and useful work.

GitHub ↗ LinkedIn ↗ Email ↗

I turn hard problems into clear products.