AI scales the decision
you already make.
First, make it the right one.
Kendall Works is an operating partner to companies that move goods and money at scale. We find the constraint, fix it where the work actually happens, then build the AI, data and automation that keep it fixed.
We take on few engagements, and we do not discuss the ones we have.
What we take on
The shape of the problem, before it has a name
It usually arrives described as a system. It is almost never a system.
Cash that is somewhere inside the business, and not in the account.
Demand visible only as far as the end of the week.
Margin that leaves quietly, somewhere between the order and the invoice.
A decision made a thousand times a day, each time slightly wrong.
Work finished weeks ago that is still not an invoice.
A process that lives, undocumented, in one person’s head.
How we work
Impact is defined before anything gets built
One method, whether it runs four weeks or two years. Most of the value is decided in the first step — the step most programmes skip.
Find the constraint
The board goal, followed down to the one thing holding it. We instrument it before we touch it.
Solve in order
People first, then the logic and the AI, then the system that carries them.
Prove it live
One territory, one channel — held against a control until it beats today.
Make it hold
Integration, controls and monitoring, so the gain survives volume and peak.
Let it compound
Each cycle adds data and the system sharpens. We stay in the loop.
Where it has turned out to sit
It has been route to market. It has been demand, inventory and imports. It has been pricing and margin, and the long quiet gap between an order and the cash.
It has also, more than once, been none of those — and finding that out early was the whole of the value. We do not keep a list of things we do. Bring the problem; we will tell you quickly whether we are the right people for it.
Surfaces we have had to read
- ERP
- POS & sell-out
- WMS & warehouse
- telematics
- bank & payment rails
- supplier EDI
- field-service logs
- price feeds
- CRM
AI we have had to build
- demand forecasting
- price elasticity
- reorder & safety stock
- route optimisation
- anomaly detection
- document extraction
- reconciliation matching
- propensity
Not a menu — just what the work has needed so far.
Our principles
Built by operators, run on evidence
We work with companies that move goods and money at scale. The method is the same every time: find the number that matters, instrument it, fix it where the work actually happens, then let the data compound.
- Operator-led. We have run operations ourselves, not only advised on them.
- Hands on the work. We design the fix and then build it — AI, pipelines and interfaces shaped to your specific constraint, not an off-the-shelf platform that needs a year of implementation. If we cannot build it, we do not recommend it.
- Evidence over opinion. Every claim gets a baseline, a control and a number. If it does not beat the way you work today, we say so.
- Start small. Fixed-scope diagnostics, so the first commitment is a cheap one.
- We say no. When the data does not support the project, we tell you.
- Discretion. We do not name clients or publish their numbers. The work is theirs; keeping quiet about it is part of the work.
Our view
Begin with the AI and you spend the year looking for a problem to point it at.