CAREER PORTFOLIO · BVLGARI AI TRANSFORMATION
01 / 11
SERGIO
PORCELLI
PORCELLI
PRODUCT MANAGER
DIGITAL
PRODUCTS
PRODUCTS
ENGINEER BY TRAINING · BUILDS WITH AI
0+
years building digital products, from QLASH to the Italian Tennis and Padel Federation. I write the requirements, run the vendors who build, and own whether it gets used.
0 SELECTED WORKS
2026
PROFILE
02 / 11
ABOUT
ME
ME
Two engineering degrees, Computer Science and Management. That is why I sit between business and technology instead of on one side of it.
ROME · ITALIAN & ENGLISH
WHAT I DO
Business needs into requirements
Delivery with external vendors
Adoption and training
Governance and go-live calls
Data analysis and KPIs
Prototyping, hands on
EDUCATION
Computer & Automation Eng.
BSc · Politecnico di Bari
Management Engineering
MSc · Politecnico di Bari
Product Management
Master · Product Heroes
TOOLBELT
Claude Code
Replit
Figma
TypeScript
SQL
Gen video
CURRENT ROLE · PRODUCT MANAGER, FITP · 2025 - TODAY
03 / 11
THREE PRODUCTS. ZERO IN-HOUSE DEVS.
SuperTennis+
Subscription streaming, mobile to TV. Rebuilt with an external vendor.
Internazionali BNL d'Italia
Fan app for the tournament. Ticketing, live data, on-site features.
myFITP
Federal app. Native tournament section and the official esports circuit.
Everything is built by vendors. My job is the interface: requirements, reviews, delivery quality, adoption.
LEGAL · PROCUREMENT · IT SYSTEMS · MARKETING
CASE STUDY · SUPERTENNIS+ · 2025-2026
04 / 11
THE BRIEF
Rebuild the Federation's streaming service as a subscription product.
WHY IT WAS HARD
One external vendor building it. Identity, membership and payments owned by three different internal groups. And an existing subscriber base nobody had planned a migration for.
0
DEVICE PLATFORMS
0
INTERNAL SYSTEM OWNERS
0
DESIGN REVIEW SESSIONS
1
LAUNCH I STOPPED
CASE STUDY · SUPERTENNIS+ · WHAT I DID
05 / 11
Mapped the real user cases, not the happy path.
THE CALL
Days before go-live, membership validation still accepted values that were not real memberships. I stopped the launch.
SCOPE I OWNED
Requirements
User case mapping
Single sign-on
Membership webhook
Subscription flows
Design review
Privacy & T&C
Store release
Go-live governance
LIVE NOW
plus.supertennis.tv ↗
CASE STUDY · myFITP · ADOPTION
06 / 11
The app shipped fine. Nobody used it.
Members kept registering on paper. The peripheral offices did not treat it as the real system of record.
DAILY ACTIVE USERS
0K
UP FROM UNDER 5,000
STAFF COMPETENCE
+0%
AFTER 3 TRAINING SESSIONS I RAN
THE FOUR LEVERS
01Structured user feedback
02A memo making it canonical
03Cross-department training
04Incentives aligned to use
In an organisation like this, digital innovation is not a technology problem. Half my time goes to change management, and that half decides the other.
CASE STUDY · QLASH · PLATFORM PRODUCT & CLUB PARTNERSHIPS
07 / 11
The platform's sponsor-facing tournament flow
QLASH COMMUNITY APP · 2021-2023
Took a tournament platform from nothing to a live product.
As product manager, not as an engineer: requirements for game integrations across mobile, PC and console, and the API and webhook spec for results, matchmaking and tournament automation.
0K+
REGISTERED USERS
0+
TOURNAMENTS RUN
0
PLATFORM FAMILIES
SOURCE ↗
AC Milan QLASH, official roster
Inter QLASH, competitive roster
Milan Games Week, with Konami
ALSO 0-TO-1
TipKlub, Mindset Lab. Concept to launch with an external agency. Past 20,000 users in two months.
TRACK RECORD · 2017-2026
08 / 11
ROLES HELD
Product Manager · FITP2025 →
Product Manager · Mindset Lab2024-25
Head of Gaming Operations · QLASH2022-25
Esport Manager · Milan & Inter2020-23
Partnerships & Community · QLASH2017-22
WHO I HAVE HAD TO CONVINCE
C-level and marketing at football clubs, motorsport brands, game publishers, consumer electronics and media. Negotiating scope with people whose incentives are not mine is most of what this job has been.
0+
PARTNERSHIPS
0+
LIVE EVENTS
0%
TEACHING RATING
BRANDS WORKED WITH


















SELF-MADE · BUILT AND SHIPPED, ON MY OWN TIME
09 / 11
I prototype the thing instead of describing it. All three are live, and none of them were briefs.
AstaCerta
astacerta.it ↗
Problem. Fantasy football leagues pick their auction date in a WhatsApp thread, badly.
Built. A scheduling SaaS with a paid tier, sold through redemption codes.
How. TypeScript monorepo, OpenAPI contract with generated client, Postgres, end-to-end suite as the release gate, EU hosting.
CalcoRALore
LIVE ON VERCEL ↗
Problem. Italian gross-to-net is opaque, and every online calculator hides its assumptions.
Built. A salary estimator where every number can be traced back to the law that produced it.
0
MUNICIPAL RATES
0
REGIONS
0
MODEL TESTS
Rates generated from the Ministry of Finance registers, each carrying its own deliberation and date. Same discipline I apply at work: cite the source, or do not state the number.
SuperTennis+ migration landing
st-plus-landing.vercel.app ↗
Problem. The vendor's landing for the platform migration was not what we had agreed, and the discussion was going in circles.
Built. A working version of the page I wanted, deployed, so the conversation had something concrete in it.
Result. Faster to build the page than to keep describing it. This is the habit the rest of this deck is really about.
HOW I WORK · TOOLS FOR MY OWN DESK
10 / 11
ASSISTANTS I BUILT FOR MY FUNCTION
One person covering strategy, product, vendors, procurement and legal. The bottleneck was never the decisions, it was the reading and drafting around them.
So I built assistants that draft briefs, review tender documents against known criteria, and turn meetings into decisions and owners. Grounded on curated internal knowledge instead of model memory, each answer carrying its source, and nothing leaves without me.
So I built assistants that draft briefs, review tender documents against known criteria, and turn meetings into decisions and owners. Grounded on curated internal knowledge instead of model memory, each answer carrying its source, and nothing leaves without me.
THE RULES THAT MADE THEM STICK
01Answer from curated knowledge, never from memory
02Always return the source
03Draft, never final
04Ask one question before producing
05Never send anything on its own
WHAT THAT TAUGHT ME
Getting something to work once is easy. Getting people to open it again six months later, under deadline, is the actual problem worth solving.
CONTACT
11 / 11