Skip to main content

Track record

What I have run

Seven years in operations: retail, B2B e-commerce, consulting. Below is the full path, plus the projects where you can check the result yourself.

The path

Four companies, the same question every time: why does this team take thirty days to do something that should take three?

  1. In's Mercato

    OperationsRetail

    Where I learned what happens when a process is not written down: in retail, the deadline does not move.

    Rules and flows made repeatable across stores and warehouse

  2. Aldi

    OperationsRetail

    A chain with high volumes and thin margins: here, two points of efficiency is the entire margin.

    Process automation, +30% stated efficiency

  3. Accenture

    Delivery & AutomationConsulting

    First client projects, with the stated goal of removing manual steps and waiting time.

    Automated processes, +30% stated efficiency

  4. Esse Solutions

    Head of OperationsB2B e-commerce

    A €500K portfolio across Magento, Shopware and Pimcore, with distributed teams of 8-12. Here I introduced Agile to an organisation that did not use it.

    +25% productivity, -40% time-to-market

Client project

One project, told in full. It is the only one where you can check my work yourself.

Panificio da Sergio

E-commerce for a family bakery, delivered and still online

Nobody asked me for a website. They had a recipe archive handwritten on paper, three generations who disagreed about everything, and a shop that ended each day not knowing what it had sold. The real work was deciding what to automate and what to leave to a human.

The problem

Three generations with incompatible opinions on pricing, catalogue and opening hours. No data on what actually sold, because it was recorded by hand. A website with a £0 materials budget.

How I handled it

  • A workshop with all three generations together, so principles were decided before specs
  • Specs written in Italian and English, so nobody could claim they had not understood
  • Open source and free tiers only: the budget constraint became the design constraint
  • A catalogue with collection scheduling, because for bread the date matters more than the price

The result

  • A complete e-commerce site in production, under the materials budget
  • End-to-end order flow: catalogue, cart, collection, confirmation, dashboard
  • Bilingual, with care for the shop's walk-in customers
  • Still online months later, which is the only test that counts
Next.js, Tailwind, TypeScript, PostgreSQL, next-intl, Netlify

What I build to do the job better

Three projects of mine, two open. Not listed to pad a CV: they are the tools behind the work above, and what they taught me.

ShipKit

A product built on the same rule I use in consulting: scope lives in a document, and anything not written down will not be done.

VulnClaw

An open source CLI for penetration testing. It started because I wanted to understand how an attack works without depending on tools I could not inspect.

It started from one scope decision: over six months I said no to 60% of requested features to ship the 40% that mattered.

github.com/Riccardobozzato94

Synapse

My knowledge vault with semantic search and memory shared between agents. Everything I learn about processes ends up here.

9,000+ notes indexed, each with an embedding and a link

It is where everything I learn about processes ends up. If I do not write it there, I forget it, and what I forget is the definition of tribal knowledge.