Technology consulting services for SMEs
Tell me where you're stuck, I look at how you work, and I tell you whether I can fix it or you're better off with someone else. That's what the first call is for.
How does the work run, and how far does my responsibility go?
Every engagement starts with a 30-minute call: you walk me through the context and I tell you whether I can be useful. Then I look at the processes, the systems already in place and who uses them, before writing a line of code.
Roadmap and prototype don't go in sequence, they move together: priorities get decided while a working prototype tests them on real data. I work from the inside, next to the people doing the job, as a Forward Deployed Engineer, and my responsibility doesn't end at release: if the system isn't adopted, I stay on it until it is.
DATA FOUNDATIONS
If the numbers don't line up, I start from the data you already have; if the server is at end of life, from where it lives. Which of the two you're dealing with, I'll tell you on the first call.
How I connect the data, what you get, what it costsDATA ENGINEERING
I build the place where your company's information becomes one source: collecting data from the systems you already run (ERP, CRM, e-commerce, the spreadsheets nobody dares touch), connecting it and keeping it current without manual work. It is not a bigger data warehouse: it is the end of silos and of the hour spent working out which file is the good one.
- Every system has its own version of the same numbers and nobody can say which one is right.
- Data gets moved by hand, in a spreadsheet, on the first of the month.
- Nobody knows where a number came from, so at the first doubt nobody trusts the report.
CLOUD DATA PLATFORM
The place your data sits in, sized for the systems you will add and not only for the ones you run today: who accesses what decided once, every number traceable back to its source, and a cloud bill you can read before it arrives.
- Every time you add a system, connecting it is a project from scratch.
- The server your systems run on is at end of life and you have been told to move everything to the cloud.
- You have to give a customer or an auditor access to part of your data, and today you can't do it without handing over everything.
DASHBOARDS AND REPORTING
The numbers you decide on, where you need them and refreshed on their own: dashboards built on the single source rather than on parallel extracts, with each metric defined once and the same for everyone. It is worth little without the platform underneath, which is why I don't sell it on its own: first the source is put in order, then you look at it.
- The report you look at in meetings is prepared by hand, and it lands after the decision is made.
- You have a dashboard nobody opens, because its numbers disagree with the ERP.
- To find out how the month went you have to ask three different people.
SOFTWARE RESCUE
When nobody dares touch a system anymore, I put it back in order or migrate it piece by piece, with any downtime agreed up front.
How it works, what you get, what it costsLEGACY SYSTEM MIGRATION
Unmanageable software left behind by a supplier who vanished, or an old system on technology nobody's trained on anymore (COBOL, unmaintained PHP, RPG on AS/400, Delphi, VB6)? I run an honest assessment and put it back in order or migrate it with an incremental approach, without starting from scratch and with any downtime agreed up front. Same goes for the management software or CRM (TeamSystem, Zucchetti, Salesforce, Dynamics, JD Edwards and beyond) you've been stuck with for years without even knowing how to read your own data. How it works, what I hand over and what it costs are on Software Rescue.
- The software was written by someone who is long gone, and nobody dares touch it.
- It runs on technology nobody knows anymore: COBOL, AS/400, ten-year-old PHP.
- The app you built with an AI tool only works when you are the one using it.
AI INTEGRATIONS
From the first use case to a system that runs every day, not a demo that impresses once.
Where to start, agents or automations, a stalled pilotAI CONSULTING FOR SMES
In 2025 only 15.7% of Italian SMEs used at least one AI technology, against 53.1% of large companies (Istat, Imprese e ICT, December 2025): not having a structured AI project yet is the norm, not a delay. I map the shadow AI already in use among the people doing the work (ChatGPT, Copilot) and help you pick the first use case worth starting with, with a pragmatic policy, not a 40-page manual.
- Everyone is talking about AI and you don't know which problem to start from.
- Your people already use ChatGPT and nobody has set any rules.
- You have seen demos that work, none of them on your actual work.
FROM PILOT TO PRODUCTION
95% of generative AI pilots produce no measurable return (MIT NANDA, The GenAI Divide, 2025): almost always because nobody stays accountable for the jump to production. As a Forward Deployed Engineer I stay in the field through the first month of live running on real data, and where the result can be measured I tie part of the fee to measured value rather than to billed time.
- You paid for a prototype that has been sitting where it was for months.
- It works on sample data, not on the real thing.
- Nobody inside owns getting it to production, so that is where it stays.
AI AGENTS & AUTOMATION
AI agents that reason, plan and act inside your processes, from qualifying leads to reading documents and case files, with human oversight by design. The goal, put plainly, is to take manual, repetitive work away from whoever does it today rather than add to it, and move those hours onto work that genuinely creates value.
- Your sales team spends more time qualifying leads and writing follow-ups than selling.
- Every week you handle the same exceptions manually.
- You would automate, but you don't trust a system that decides on its own.
SECURITY AND COMPLIANCE
What you actually risk, what your client is asking for and where to start, without selling fear.
What you risk, what I do, what it costsCYBERSECURITY FOR SMES
Your data and your systems are not secured once the project is finished: the perimeter is built while I put them into production. Ransomware and data breaches increasingly hit SMEs, often the weakest link in a larger client's supply chain. I run an honest assessment of the real risk, from basic security posture to NIS2 compliance, without selling fear or solutions that never touch the underlying problem.
- If your systems were locked tomorrow, you don't know what stops or for how long.
- A large client is asking for security requirements you have no way to meet.
- The backups exist, but nobody has ever tried restoring them.
CUSTOM TOOLS AND GOVERNANCE
The tools you are missing and the calls you keep postponing: the costliest one is the decision nobody makes.
What I build, at what cadence, what it costsWEB & MOBILE APP DEVELOPMENT
I build web apps, mobile apps and custom tools, integrated with the systems you already have, from MVP to production: I assess feasibility, not sell a standard package. Data & AI native, with technical SEO and GEO handled from the first commit.
- You are holding a process together with a spreadsheet when it now needs a real tool.
- You were quoted a standard package that doesn't solve your case.
- You need an application, and it has to talk to the systems you already run.
FRACTIONAL CTO / CDO / CIO
As your Fractional CTO / CDO / CIO I step in for platform choice, vendor contracts, architecture review. Same hands-on Forward DeployedEngineer approach, with a cadence agreed on what you actually need: some companies need a day a week, others half a day a month, without the cost of a full-time executive. The same applies to AI governance: if you need the Chief AI Officer function without hiring one full-time, I cover it under the same fractional model.
- You have to pick a platform and nobody inside can tell you if it is the right one.
- You sign contracts with technical vendors with no way to judge what is in them.
- Whoever handles the technical side has nobody more experienced to answer to.
TRAINING
When you don't want to buy the work but to learn how to do it: three programs for your own people, on your own data and processes.
How a day works, what it costs, why no certificateAGENTIC AI ADOPTION
courseA path to adopt Agentic AI in your business: from autonomous agents that reason, plan and act to intelligent process automation.
- Fundamentals: LLMs, advanced prompt engineering and multi-agent architectures.
- Orchestration: how to make multiple agents collaborate with each other and with your existing tools, whatever the stack.
- Use cases: Agents for document analysis, decision support, workflow automation and customer operations.
DevOps & CLOUD ENGINEERING
courseHow to ship software without dreading Friday afternoon: DevOps practices applied to a small team rather than to a multinational.
- Infrastructure as Code: declarative, reproducible management of cloud resources, whatever the provider, so nothing is configured by hand.
- CI/CD and safe releases: automated tests, separate environments, and the ability to roll back in five minutes instead of overnight.
- Observability and GitOps: monitoring, alerting and versioning of what runs, so when something breaks you know where to look.
DATA STORYTELLING
courseCommunicating with data is strategic. You'll learn to turn complex analyses into persuasive and actionable narratives.
- Visual design: Principles of visual perception, chart selection and accessibility.
- Tools: whatever BI or data viz platform your company already uses, adapted for presentations that are genuinely data-driven.
- Storytelling framework: Narrative structure, context, insight and call-to-action for every audience.