AI AUTOMATION · LOCAL AI · SMART SYSTEMS

I’m Marco Figaro.
I design AI tools.

I build assistants and automations that organise information, connect tools and make every action verifiable.

25+
Years of IT experienceTechnical support, systems and digital training.

NEW PROJECT · IN DEVELOPMENT

CharMe

A companion that starts with you.

A one-of-a-kind companion, with a personality tailored by you. Or a digital clone of yourself or a remembrance companion inspired by someone dear to you, shaped from the materials you provide.

Coming soon.
01Local AI02Automation03RAG04Multimodal systems05Apple ecosystem06Privacy by design

01 / SELECTED WORK

Selected work

Concrete systems, one recurring obsession:
making AI useful, legible and controllable.

CASE STUDY01

PERSONAL AI · APPLE ECOSYSTEM

Cassandra

Cassandra brings conversation, persistent memory, Organizer, document search and scheduled tasks together in a personal Mac app. Cloud AI interprets requests; the local bridge manages data, tools and verification. Apple integrations connect the workspace to voice and smart home controls, keeping actions under control.

My work on the project

Problem
Bring information, reminders and tasks together in a personal assistant with controlled actions.
My contribution
I designed the desktop app, memory and Organizer, connecting local tools and Apple integrations.
What you can see
The overview illustrates the path from request to action and verification. Real recordings are coming soon.
Flagship projectDesktop app for MacCloud AI · local workspace
  • ✦Conversation and persistent memory
  • ✦Notes, events and reminders in Organizer
  • ✦Document search, jobs and local skills
  • ✦Apple integrations and controlled actions
Preview Cassandra
Mac · Memory · Organizer

02 / ABOUT

Marco Figaro.
Applied AI.

I design AI assistants, automations and local systems that turn concrete problems into clear, verifiable tools for people.

I bring more than twenty-five years of IT, technical support and digital training experience into documented workflows: understandable architectures, explicit limits and controllable results.

03 / METHOD

Building with
intention.

An automated answer is not enough.
It must be possible to understand why.

01

Understand the context

Documents, constraints, people and data before technology.

02

Make the flow legible

Roles, sources, state and errors visible to the people using the system.

03

Verify the result

Tests, audits and a clear distinction between execution and evidence.

ROLE

Concept, architecture and end-to-end development

STACK

Electron · Apple Shortcuts · HomeKit · Ollama Cloud · SQLite

STATUS

Working prototype in evolution

CONTRIBUTION

Desktop app, memory and Organizer, local bridge, Apple integrations and result verification.

Email ↗LinkedIn ↗GitHub ↗