Automation and AI that make it to production.
I help large organisations work out where automation and AI create value, and I stay with them through to release. Twelve years of enterprise delivery, an engineer’s method, a product owner’s craft.
I’m Gianluca. Computer engineer from Politecnico di Milano, certified Product Owner, independent since 2018 and based in Milan.
Over twelve years I have taken complex projects into production for large organisations — retail, financial services, pharma, automotive, logistics — coordinating remote teams of up to sixteen people. Since 2018 I have worked on applied AI: first voice interfaces, then conversational automation, today automation and agents on enterprise platforms.
I work alone but not in isolation: I coordinate the client’s technical teams, or assemble the people the job needs. And when a use case does not hold up, I say so before we start — not halfway through.
Five phases, from problem to release.
I don’t come in as a technology vendor. I come in earlier — to work out what is worth doing — and I stay until it is in production and being measured.
Explore
- Envisioning workshop
- Industry scenarios
- Opportunities
- Priorities
Understand
- Interviews
- Stakeholders
- Process mapping
- Requirements
Design
- Solution design
- Functional architecture
- Figma prototype
Build
- Working prototype
- MVP with AI tooling
- User validation
Deliver
- Backlog and sprints
- Prioritisation
- Release
- Measurement
Method, prototype, delivery.
Strategy consultants stop at the slides. Engineers can build, but they don’t facilitate workshops, run discovery or manage stakeholders. I cover the whole arc.
Method
Envisioning workshops, interviews, requirements, process mapping. Before building anything, I make sure it is the right thing to build.
Prototype
Solution design and a Figma prototype, or a working MVP built with AI tooling. You can see it, try it and judge it — before investing in development.
Delivery
Backlog, sprints, coordinated teams, priorities set by real value. Through to production release and measurement.
Agents are a continuation, not a change of field.
I have worked on applied AI since 2018: voice skills for pharma, an omnichannel conversational AI platform before generative AI reached the market, the product side of an AI-focused startup studio, and today automation and agents on enterprise platforms.
- 2018Voice AIAlexa and Google Home skills for pharma
- 2019Conversational AIOmnichannel platform, before generative AI
- 2021AI productAI-focused startup studio
- 2024Automation and agentsBoomi, agents included, and Shopify
- 2026Agentic AIMIT Sloan and an L0–L5 build track
Where I have already worked.
- Retail and omnichannelStore networks, e-commerce, channel integration
- Financial servicesBanking, wealth management, regulatory programmes
- Pharma and healthcareMedical affairs, operations, patient services
- AutomotiveDealer networks, plant monitoring, lead generation
- Logistics and servicesOperational processes and workflow automation
- Innovation and educationTraining programmes, open innovation, startup mentorship
EssilorLuxottica · MediaWorld · Poste Italiane · Banca Mediolanum · Fideuram · Intesa Sanpaolo · Janssen · Arcese · MotorK · Talent Garden
What the people who worked with me say.
«A safe pair of hands in a world of talkers.»
He is an engineer who rejected the dogmas and took the “it can be done” route. Attentive to detail. Direct. Flexible. He chooses who he works with, because he does not want to work with everyone.
Extremely versatile and well-rounded. He adapts to the situation and is effective with industrial clients and individual startups alike. Very concrete, a simplifier, he gets to the heart of the matter.
Engineers and designers are said to speak different languages, but I really valued the way Gianluca built a working relationship across our multidisciplinary team as a manager. He always found time for teaching, coaching us through feedback on our work.
Public LinkedIn recommendations, quoted in full and translated from Italian.
Got a use case in mind — or just a suspicion there is one?
The first step is a conversation. If there is nothing worth building, I will say so: that is the part of the job that saves the most money.