Arcane Intelligence

Arcane Intelligence

Arcane Intelligence

Writing

The Loop

Bring one decision

Bring one decision

Building Intelligent Organisations

AI made your organisation faster.
Not better at deciding.

AI made your organisation faster.
Not better at deciding.

I work with chief executives to find where a critical decision breaks and rebuild the intelligence behind it.

I work with chief executives to find where a critical decision breaks and rebuild the intelligence behind it.

Bring one decision

Former CDTO, Unilever International

Former CDTO, Unilever International

Former CDTO, Unilever International

€2B business · 100+ markets · Visiting Scholar, Ben-Gurion University

€2B business · 100+ markets · Visiting Scholar, Ben-Gurion University

€2B business · 100+ markets · Visiting Scholar, Ben-Gurion University

What I do today

Advisory to a Fortune 500 leadership team on leading in the AI age.

Advisory to a Fortune 500 leadership team on leading in the AI age.

Fractional CDTO to privately held groups.

Fractional CDTO to privately held groups.

Board advisor to startups in data, analytics, AI and defence technology.

Board advisor to startups in data, analytics, AI and defence technology.

I don’t name clients.

The intelligence loop

Speed is not intelligence.

Speed is not intelligence.

AI doesn’t make an organisation intelligent.

It makes it faster at whatever it already does.

AI doesn’t make an organisation intelligent.

It makes it faster at whatever it already does.

PURPOSE

SENSE

UNDERSTAND

DECIDE

ACT

LEARN

PURPOSE

SENSE

UNDERSTAND

DECIDE

ACT

LEARN

Every organisation already runs this loop.
The question is what it runs for, and where it breaks.

It rarely breaks in one place. The symptom shows up far from the cause. That is why tracing the whole cycle is the work.

It rarely breaks in one place. The symptom shows up far from the cause. That is why tracing the whole cycle is the work.

Intelligence is a property of the organisation, not of the tools inside it.

Intelligence is a property of the organisation, not of the tools inside it.

Below: three loops, and which stages each one rebuilt.

Three loops, rebuilt · Unilever International

$10M+ in orders recovered.

$10M+ in orders recovered.

$10M+ in orders recovered.

Decision. Should we accept this order at this price?

Before. Information sat across markets and systems. A manual check took ten hours, repeated market by market, so orders were missed or lost.

After. An agentic platform connected the information and the logic, so every order in 100+ markets could be checked the same way in two minutes.

Consequence. $10M+ in orders recovered every year. Every order checked, not a sample. Manual effort down 95%.

Decision. Should we accept this order at this price?

Before. Information sat across markets and systems. A manual check took ten hours, repeated market by market, so orders were missed or lost.

After. An agentic platform connected the information and the logic, so every order in 100+ markets could be checked the same way in two minutes.

Consequence. $10M+ in orders recovered every year. Every order checked, not a sample. Manual effort down 95%.

Decision. Should we accept this order at this price?

Before. Information sat across markets and systems. A manual check took ten hours, repeated market by market, so orders were missed or lost.

After. An agentic platform connected the information and the logic, so every order in 100+ markets could be checked the same way in two minutes.

Consequence. $10M+ in orders recovered every year. Every order checked, not a sample. Manual effort down 95%.

Stages rebuilt

UNDERSTAND · DECIDE · ACT

Fill rate from 60% to 90%.

Fill rate from 60% to 90%.

Fill rate from 60% to 90%.

Decision. Which distributors should we back, with what stock, and what must we learn next?

Before. Performance was scattered across hundreds of incompatible signals. There was no common basis for a decision.

After. Sales, inventory and demand signals were integrated into one view. AI forecast the demand, so teams could act and learn from the outcome.

Consequence. Order fill rate moved from 60% to 90%+. Forecast accuracy reached 70%. Planning cycle 90% shorter.

Decision. Which distributors should we back, with what stock, and what must we learn next?

Before. Performance was scattered across hundreds of incompatible signals. There was no common basis for a decision.

After. Sales, inventory and demand signals were integrated into one view. AI forecast the demand, so teams could act and learn from the outcome.

Consequence. Order fill rate moved from 60% to 90%+. Forecast accuracy reached 70%. Planning cycle 90% shorter.

Decision. Which distributors should we back, with what stock, and what must we learn next?

Before. Performance was scattered across hundreds of incompatible signals. There was no common basis for a decision.

After. Sales, inventory and demand signals were integrated into one view. AI forecast the demand, so teams could act and learn from the outcome.

Consequence. Order fill rate moved from 60% to 90%+. Forecast accuracy reached 70%. Planning cycle 90% shorter.

Stages rebuilt

SENSE · UNDERSTAND · DECIDE · LEARN

NPS from −60 to +40.

NPS from −60 to +40.

NPS from −60 to +40.

Decision. What do our customers need, and what should we change because of it?

Before. Every order and answer depended on a person. 1,000+ trade customers waited on someone else’s availability.

After. Digital marketing, connected commerce and self-service captured customer signals, turned them into action, and measured the response.

Consequence. Net Promoter Score moved from −60 to +40 in two years, across 100+ markets. Service scaled three times over without adding headcount.

Decision. What do our customers need, and what should we change because of it?

Before. Every order and answer depended on a person. 1,000+ trade customers waited on someone else’s availability.

After. Digital marketing, connected commerce and self-service captured customer signals, turned them into action, and measured the response.

Consequence. Net Promoter Score moved from −60 to +40 in two years, across 100+ markets. Service scaled three times over without adding headcount.

Decision. What do our customers need, and what should we change because of it?

Before. Every order and answer depended on a person. 1,000+ trade customers waited on someone else’s availability.

After. Digital marketing, connected commerce and self-service captured customer signals, turned them into action, and measured the response.

Consequence. Net Promoter Score moved from −60 to +40 in two years, across 100+ markets. Service scaled three times over without adding headcount.

Stages rebuilt

SENSE · ACT · LEARN

“Dany is that rare leader who possesses both high-level strategic vision and a sharp focus on operational detail.”

“Dany is that rare leader who possesses both high-level strategic vision and a sharp focus on operational detail.”

— RK Chaudhari, Global Customer Development Strategy & Operations Leader, Unilever International

“He has a gift for understanding complex business issues and then quickly turning them around into easily executable business solutions.”

“He has a gift for understanding complex business issues and then quickly turning them around into easily executable business solutions.”

— Vikram Godboley, General Manager, Unilever International · Worked with Dany for 15+ years

Dany Krivoshey

DESCRIBE BEFORE YOU PRESCRIBE.

My father has practised medicine for 55 years. He listens, gathers evidence, prescribes, then watches what happens.

The principle has stayed with me: understand what is happening before deciding what to change.

Organisations are no different. A visible problem is often only the symptom. The real work is finding the cause.

Twenty years at Unilever International, latterly as Chief Digital & Technology Officer across 100+ markets, taught me that technology is rarely the hard part.

I help CEOs turn complexity into clarity—understanding what matters, making informed decisions, acting on them, and learning from the results.

That is how we build intelligent organisations.

Dany Krivoshey

Founder, Arcane Intelligence · Former CDTO, Unilever International

Visiting Scholar, Ben-Gurion University

What you buy

One critical decision → one diagnosed failure →
one rebuilt operating loop → one measurable result.

Start here

01

Loop Diagnostic

2 weeks

One decision, traced end to end. Where it breaks, and what intelligence would actually change.

Loop map · three prioritised fixes · executive readout

01

Loop Diagnostic

2 weeks

One decision, traced end to end. Where it breaks, and what intelligence would actually change.

Loop map · three prioritised fixes · executive readout

01

Loop Diagnostic

2 weeks

One decision, traced end to end. Where it breaks, and what intelligence would actually change.

Loop map · three prioritised fixes · executive readout

Bring one decision

02

Loop Rebuild

4–6 weeks

The loop redesigned, governed proportionately, and one use case live with the people ready to move.

Redesigned loop · live use case · before / after metric

02

Loop Rebuild

4–6 weeks

The loop redesigned, governed proportionately, and one use case live with the people ready to move.

Redesigned loop · live use case · before / after metric

03

Executive Advisor

Monthly

A sparring partner for you and the board while the loop keeps learning.

Board sessions · decision reviews · on-call counsel

03

Executive Advisor

Monthly

A sparring partner for you and the board while the loop keeps learning.

Board sessions · decision reviews · on-call counsel

04

Fractional CDTO

At capacity

I hold the seat: strategy, team, roadmap and delivery, at the seniority the role needs. No full-time executive on the payroll.

Currently full · enquiries taken for future availability

04

Fractional CDTO

At capacity

I hold the seat: strategy, team, roadmap and delivery, at the seniority the role needs. No full-time executive on the payroll.

Currently full · enquiries taken for future availability

A handful of relationships a year.

A handful of relationships a year.

Bring one decision you don’t trust, and we’ll find where it breaks and whether it is worth rebuilding.

Bring one decision

Arcane Intelligence

Arcane: known to few. That’s the problem.

Connect

Connect. Understand. Evolve.

Connect. Understand. Evolve.

LinkedIn →

LinkedIn →

Website

arcaneintelligence.ai