"The same supply chain discipline that drives the world's best companies should be available to any company that needs it. AI makes that possible."

Why this exists

Supply chains are the backbone of global commerce. Most companies run them with overworked internal teams or expensive consultants who leave. AI changes this equation, but only if it is built into the operating model from the start.

Traditional consulting models are too slow and too expensive. SaaS tools sell licenses but do not operate. There was no firm that combined deep supply chain expertise with purpose-built AI and stayed on as an operating partner.

AI agent technology has reached the point where purpose-built agents can handle the analytical volume that previously required large teams. The companies that adopt this model first will have a structural advantage. Williams was built to be that firm.

Mauro Cozzi, Founder and CEO

Mauro has spent a decade building supply chain solutions across strategy, technology, and operations at companies ranging from $500M to $50B in revenue. He has led planning transformations, built AI-powered forecasting systems, and managed cross-functional supply chain programs spanning procurement, logistics, and S&OP.

He saw the same pattern at every company: consultants deliver great analysis, leave a deck, and nothing changes. Internal teams are too stretched to execute. The gap is not insight. It is operational capacity. Williams was built to fill that gap with AI-native operations that actually run, not just advise.

The team behind the technology

25+
Years of combined supply chain leadership
$10B+
Supply chain spend managed across careers
4
Industries with deep operational experience
PhD
AI research team with published work at NeurIPS, ICML

"We built Williams because we saw both sides of the problem. Supply chain leaders who understand the domain but lack AI capabilities. AI teams who understand the math but have never run an S&OP cycle. You need both in the same room."

Two disciplines. One team. Neither works without the other.

Most AI companies hire engineers and try to learn the domain later. Most consultancies hire domain experts and try to bolt on technology. Williams was built with both from day one. Our supply chain professionals and our AI researchers sit in the same room, work on the same problems, and build solutions together.

That is why our AI actually works in practice: it was designed by people who understand both the mathematics and the messy reality of supply chain operations.

Supply chain leadership

Our front-of-house team brings decades of experience running global supply chains at scale. They have managed billions in spend, led teams of hundreds, and operated across every function from procurement to logistics to finance. They have seen what works and what does not, at every scale, in every sector.

Global CSCO experience
Former Chief Supply Chain Officer at a top-5 global FMCG company. 25+ years leading procurement, logistics, finance, and strategy across 190 countries.
Ranked #1 Supply Chain Executive in Europe (2018, 2019)
Board-level networks
Active board membership at global logistics and shipping companies. Deep relationships across the European supply chain ecosystem.
Operational depth
Our Supply Chain Partners have all held senior supply chain roles in industry. They have run S&OP cycles, negotiated with suppliers, managed inventory. They speak the language because they have done the work.
Sector expertise
FMCG, food and beverage, manufacturing, pharmaceuticals, retail. Our team has operated inside these industries, not just advised them.

AI research and engineering

Our back-of-house team is a group of AI researchers and engineers with PhDs from leading institutions. They do not just apply off-the-shelf models. They design novel architectures for the specific challenges of supply chain data: sparse, noisy, multimodal, and high-stakes.

PhD research team
AI researchers with doctorates in machine learning, information theory, and neurosymbolic systems. Published at NeurIPS, ICML, and AAAI.
Research-led AI team, building, not advising
Proprietary platform: Durable
We built Durable, an information-theoretic discovery engine that automates end-to-end workflows. It discovers signals, validates findings, synthesises code, and executes multi-step processes autonomously with human approval gates. It can automatically build integrations against any internal or external system, maintain those integrations over time, and convert complex process descriptions and natural language feedback into defined, executable workflows.
durable.ai / discover, validate, execute / full cycle automation
Agents in production, and rapid development
Three specialised agents (Sensing, Review, Outreach) running in production today. But the real advantage is what comes next: because they run on Durable, we can design and deploy a new agent for a new use case in days, not months. Quality monitoring, trade compliance, contract analysis, whatever the problem requires.
Agents in production / new agents deployable in days, not months
Supply chain-native AI
Every model is designed for supply chain operations specifically. We handle intermittent demand, promotional spikes, supplier variability, and long-tail SKU distributions, problems generic AI tools struggle with.

AI without domain expertise is dangerous.
Domain expertise without AI is slow.

The result: you get world-class supply chain execution capability and strategy across the board, without building a 50-person team to do it.

25+
Years of supply chain leadership at global scale
190
Countries our leadership team has operated across
Days
To deploy a new AI agent for any use case
London
Headquartered, with team and clients across the UK, Europe and North America

Experience built on doing, not advising

Williams hires people who have actually run supply chains, not just consulted on them. Our team combines operational supply chain veterans with AI engineers. We look for people who have made real decisions under pressure, lived with the consequences, and learned from both successes and failures. That is what enables us to build systems that work in practice, not just in theory.

Principles we hold ourselves to

Outcomes over activity. We do not measure success by hours worked or reports generated. We measure it by the documented impact on your KPIs. If we have not moved the needle, we have not delivered.

Transparency over polish. We show you exactly what our systems are doing, how they reach their conclusions, and where they are uncertain. No black boxes. No hand-waving.

Deliberate growth. We add clients only when we are confident we can deliver outcomes. We would rather do excellent work for fewer clients than mediocre work for many.

Start small, prove value. Every engagement begins with a diagnostic and a defined pilot. We earn the right to expand by delivering results on the first workstream, not by selling a vision.

Your team, not a replacement. We work alongside your people, not instead of them. Our goal is to make your supply chain function stronger, whether we are managing it or training your team to use our tools.

Skin in the game. Our pricing is tied to outcomes. We only earn our performance fee when we deliver documented, measurable results that your team can verify independently.

Advisory network. Our advisory network is being built with senior supply chain and AI leaders. We will share more as appointments are confirmed.

Want to meet the team?

Book a consultation. We will bring the right people to the conversation based on your priorities.

Book a consultation
hello@williams.supply