Digital-asset benchmarks & market data
Designed a scalable crypto-asset benchmark and the infrastructure behind it, connecting market-data APIs and building ETL workflows for high-frequency data.
I'm Marcello Pichini, a product leader bringing together digital assets, AI, and financial data.
At SIX Group, I work on crypto indices and financial data products, owning the roadmap for a broader portfolio of 300+ indices across crypto, global, Swiss, and ESG markets. I connect product strategy with quantitative research, engineering, and reliable market data.
My digital-assets work began at CheckSig, where I designed a crypto-asset benchmark and built the market-data infrastructure behind it. At IBM, I developed NLP and deep-learning systems for financial institutions. I hold master's degrees in Data Science and FinTech, with a data science thesis on Transformers for automated question answering.
I also co-founded ./cogita, a nonprofit where we make machine learning and generative AI accessible to broader audiences. Our courses have reached more than 12,000 learners.
Designed a scalable crypto-asset benchmark and the infrastructure behind it, connecting market-data APIs and building ETL workflows for high-frequency data.
Led a platform integration, migrated more than 100 indices, and scaled the production portfolio to over 300. Improved product development processes.
Built pipelines that ingest and normalize transactional and order-book data from multiple sources, with automated quality controls to validate index calculations.
Developed conversational assistants, NLP chatbots, and computer-vision proofs of concept for financial and payments institutions, from problem framing to client validation.
A catalog of scientific discoveries enabled by AI, each entry linked to its original announcement, with filters, model-family profiles and open data under CC BY 4.0.
SIX Group · Zürich, Switzerland
End-to-end ownership of 300+ indices across four product families. Product Owner for a six-person cross-functional team, connecting engineering, quantitative research, and commercial stakeholders.
IBM · Milan, Italy
Built AI/ML proofs of concept for financial-services clients and architected cloud-based data pipelines and NLP and deep-learning models.
CheckSig · Milan, Italy
Owned quantitative methodology and technical design for a crypto-asset benchmark, with responsibility for data architecture, AWS infrastructure, and testing.
As co-founder of ./cogita since 2022, I design and deliver courses in machine learning and generative AI, including introductory LLM workshops for non-technical audiences.
I lead volunteers and employees, manage budgets and resource allocation, and translate complex AI concepts into practical learning experiences.
University of Milano-Bicocca · 108/110
Thesis: Transformers for automated question answering
LUMSA University · 107/110
University of Bologna / Rotterdam University of Applied Sciences

DeepLearning.AI
Issued
Large Language Models · Generative AI

SAFe by Scaled Agile, Inc.
Issued
Expired
Product Management
NOVA 111 — Best Under 35 in Italy, Finance & Payments

A week after the cryptography essay: three security failures, and what happens when agents keep searching for the next one.
Read article ↗
AI-resistant versus quantum-resistant cryptography, and the risks that reach far beyond cryptocurrency wallets.
Read article ↗
Terafab, orbital data centres, and the physical infrastructure trying to keep up with AI.
Read article ↗
What happens when software becomes the most frequent customer? Test the payment demand scenarios and compare Bitcoin, Ethereum, Visa and Kaspa.
Read article ↗
Model releases are arriving within weeks. Explore the timeline, the benchmarks, and what a self-improving research loop could change.
Read article ↗
Could AI give us more freedom to create, learn, and follow our passions? A reflection on technology and the possibility of a new Renaissance.
Read articleFor conversations about digital assets, AI, financial data, or building better products.