Luca Mastrostefano

A crypto predictive engine leveraging real-time cross-asset prediction divergence and historical Bayesian states to capture high-conviction trading opportunities.

• Led a team of 8 in build a product to fuel a step-change in a traditional industry, rethinking first-principle how to collect a proprietary data layer to solve the biggest problem in industry: how to identify top talent.

getCredible helps companies automatically identify and hire top performers by combining peer referrals with AI.

Users play quick 1-vs-4 polls featuring AI-selected talent from their contact list (similar to Nikita Bier's Gas app) to vote for the top performers they know, and reserve a 2% success fee with a single click on the selected talent. Our AI aggregates and analyzes these peer insights into verified performance leaderboards by role, industry, and seniority.

Companies hire directly from these performance leaderboards with a single click, paying a 10-30% success fee (no upfront costs), or purchase performance scores of their applicants ($1K-$5K per job position) to optimize their hiring channels.

getCredible Pitch:

Product Screens:

• As Founder and CTO, I led a cross-functional team of ~20 to build and scale a complete B2E ecosystem (mobile app, back-office, support) from scratch in less than 3 months. We secured a €1M contract shortly after launch.
• I managed strategic relationships and decision-making directly with C-Level executives of the publicly listed partner company (Be Group - Ticker: BEST.MI).
• Our hardware solution was trust-tested for high-stakes contact tracing, notably during the Cortina 2021 World Ski Championships (the official Olympic test event) and the Serie A soccer match during the pandemic (Udinese-Fiorentina).

Human Mobility Press Release (Partnership with Be):

In just 3 months, I built and led a remote cross-functional team of 35 world-class researchers and engineers from major centers including IBM, Massachusetts Institute of Technology, University of Southern California, USFC, CNR, etc..

We developed the world's first graph-based contact tracing algorithm and mobile application, which was officially shortlisted by the Italian Government as the primary backup to the Apple/Google system, out of 350 worldwide entries, after passing rigorous national security and technical pressure tests.
View Official Report →

My work on the foundational graph principles was co-authored and published with researchers from MIT and USC in ScienceDirect: Defining Graph-based Contact Tracing.

I led the development of the intelligent core of InReach, the AI-powered investment firm focused on early-stage European startups. The automatic pre-qualification system reached 70% agreement with the partner's pre-qualification decision, based on ~100 features extracted for each company, allowing the firm to scale operations without compromising quality.

Noticeable investments: Soldo, Shapr3D, Dreamdata, Kayzen, Craft, Eneba, Oberlo (exit), Sealk (exit), Qriously (exit), Returnly (exit)

Built and sold an automated lead generation pipeline for translation requests by scraping forums, detecting relevant opportunities, and engaging users with context-aware quotes. Designed a dashboard to optimize performance and run A/B tests.

Sold after 2 months to Translated, grew their cold outreach channel from €200K/year to €2M/year.

Joined as the 3rd engineer to build the foundational infrastructure of a startup that would later reach nearly-unicorn status as a global leader in AI translation.

• Completely reproduced the entire game logic (building, resources, building constraints, etc.) on my laptop to simulate optimal strategies.
• Found the shortest path to complete mandatory Quests to unlock referral bonuses automatically.
• Still remember the terrible 2k lines of VB6 code that automatically played the game to generate referral bonuses!

• Built this game as a personal competition against my older brother to see who could create the better experience.
• Developed my very first AI (named Josha, in honor of the movie WarGames). It was a simple subtraction to make the police car chase the player's position, but it felt too cool at the time.
• I still remember friends playing so much that they eventually beat my own high score; I still smile thinking about that feeling.

I spent countless hours coding, recreating entire games like Pac-Man, Risk, and Backgammon without even knowing that arrays existed. Every object was numbered and manually updated one by one: a true obsession was born.