How long does a project take?

A web MVP starts at 4-6 weeks, a custom ERP runs 3-6 months depending on complexity. We agree clear timelines at kickoff and deliver navigable increments every 2 weeks.

Do you only work in Rome or across Europe?

We work remotely with clients across Italy and Europe. For enterprise projects we travel on-site when needed, especially during discovery and go-live.

Does the customer own the code?

Yes, always. At the end of every project we hand over the full source code, infrastructure credentials and technical documentation. No lock-in.

Do you handle maintenance after release?

Yes, we offer evolutionary and corrective maintenance contracts with dedicated SLAs. We monitor performance, security and availability.

How much does custom software cost?

It depends on scope. As a benchmark: a website starts around €8,000, an MVP from €25,000, an enterprise ERP from €60,000 upwards. We always provide a detailed quote after a free discovery call.

Rinoova Academy: training on AI, automation and MCP

A free course with a verifiable certificate, three limited-seat executive masters and tailored programmes for companies. Here's how our academy is built and who it's for.

We've launched Rinoova Academy: deep technical content on AI, automation and the Model Context Protocol, built for people who have to answer for results inside a company — not for people who collect slides.

The approach is the same one we use on client projects: start from a real problem, end up with something that works in production. Here are the three ways in.

Open courses, free, with a certificate

Open to everyone, no sign-up required. You read the material, pass the final test with at least 80% correct answers and receive a personalised Certificate of Completion, with a code and a public verification page. It's an Open Badge under the 2.0 standard: you can import it into LinkedIn and into digital skill wallets.

The first course available is Model Context Protocol: Fundamentals — roughly two hours of reading, 11 chapters, 10 examples of real protocol messages. It starts from the integration problem and ends with enterprise adoption, covering host/client/server architecture, resources, tools, prompts, transports (stdio and Streamable HTTP), OAuth 2.1 authorisation and operational security.

If you've read our piece on turning existing APIs into an MCP server, this course is the layer underneath: the why behind the how.

Executive masters for IT professionals

Deep-dive programmes with limited seats, either one-to-one with a dedicated tutor or in a small group. Content and certificate are identical in both formats; what changes is the pace. Every application is assessed individually on professional background — no degree required, but real backend development experience is.

MCP Architect — 100 hours, 4 modules. Designing, securing and operating MCP servers in enterprise environments with genuine reliability and compliance requirements. For software architects, lead developers, platform/DevOps teams and CTOs.
AI Agent Engineering — 100 hours, 7 modules. Designing agents that hold up under real workloads: tool design, context and memory management, systematic evaluation, deployment and defence against prompt injection. Beyond the proof of concept.
AI Coding Orchestration — 90 hours, 6 modules. Bringing autonomous coding agents into a real development workflow: breaking down work, multi-agent orchestration, quality checks, governance and accountability. You'll need a real codebase to work on during the programme.

Tailored corporate training

No classroom here: we work on your codebase, from day one. Coding agents such as Claude Code, Antigravity and Codex applied to your actual projects, across four areas:

Tool mastery — installing and using agentic environments day to day, context files, integration via MCP. Command-line agents change the way you work, but only if they're set up properly.
Work on your own projects — refactoring, test coverage on legacy code, framework migrations, open bugs.
Integration into your existing process — where the agent fits into the git flow, how its output gets reviewed, which metrics tell you whether it's working. An agent operating outside your process creates more problems than it solves.
Governance, security and compliance — intellectual property of generated code, traceability of changes, data usage policies.

The method: a one-hour intro call, a one- or two-day technical assessment, hands-on work alongside your developers, then handover of the skills to the team. After that you're on your own — that's the whole point.

Where to start

If you want to understand what we mean when we say MCP, the free course is right there and doesn't even ask for an email. If you're considering something more substantial, take a look at the masters or get in touch and we'll talk it through.