Custom AI and software for products that actually work.

I help teams and companies turn messy processes into practical tools: APIs, integrations, automations, and AI features connected to the data and tools they already use.

01

Scenario

Customer requests to route

Messages arrive every day and need to be read, qualified, and assigned to the right person.

Intervention

An AI assistant connected to the CRM filters requests, answers where it can, and passes only what matters to the team.

02

Scenario

Documents copied by hand

Invoices, orders, and attachments are opened one by one so data can be entered into the ERP.

Intervention

A pipeline extracts information, flags uncertainty, and syncs the validated data.

03

Scenario

Reports that take time

Data spread across different tools is collected, cleaned, and formatted every week.

Intervention

A scheduled job prepares the report, sends it to the team, and leaves only the final review.

Browse scenarios

Stack

The tools I use every day.

A pragmatic set for building, integrating, and deploying solutions that stay maintainable.

Backend & product

.NET
PostgreSQL
React
TypeScript

AI & Automation

OpenAI
Claude
RAG
Temporal

Deploy & infrastructure

Docker
GitHub
Cloudflare

About

Matteo Vannucchi

Italy / remote

I am a freelance developer in Italy. After a bachelor's degree in Computer Science and a master's degree in Artificial Intelligence in Bologna, I started working on software that removes friction: backend systems, integrations, automations, and AI features when they are actually useful.

I like starting from concrete problems, keeping theory short, and getting quickly to something you can try. Not everything needs to be automated: sometimes the right move is connecting two tools, fixing a flow, or building a small internal tool well.

First step

How we start

You do not need to arrive with a full specification. Tell me what currently takes time, creates mistakes, or blocks the work; I help you understand the simplest way to solve it, whether that is an AI feature, an integration, an automation, or a custom tool.

  1. 01

    Tell me the flow

    What happens today, which tools you use, and where data, requests, or manual decisions enter the work.

  2. 02

    Find the right point

    Not everything should be automated. I look for the small intervention that can remove friction quickly.

  3. 03

    Build something real

    A first version connected to real data or tools, tested in daily work and improved with concrete feedback.

Contact

Let's talk.

For a project, a collaboration, or a technical problem to fix, send me a few lines of context. Then we can understand whether it makes sense to start and what the simplest first step is.