Build the operational backbone. Layer AI on top.
Hand it back running.

A small, senior team that advises and builds the operating processes and the AI that runs on top of them

AI is only as good as what it runs on. We build both.

﹡What we do

Three practices.

Three disciplines, one aim: a business that runs better and decides faster. Increasingly, AI runs through all of them - the same work that makes an operation better is what makes AI worth deploying.

Operational Excellence
What we deliver

Floor-level rigor: throughput, quality, cost, lean and theory-of-constraints — with capability that outlasts the engagement.

In practice

Process redesign and standard work, increasingly with the tools and automation that make the gains bigger and easier to hold.

Supply Chain
What we deliver

End-to-end network, planning, S&OP, logistics, and footprint designed by people who've owned the cycle themselves.

In practice

Planning cadence, inventory policy, network design, and more - with intelligence and automation layered in where they compound.

AI Deployment
What we deliver

Finding where AI genuinely creates value, building the tools and agents, and making them hold up in daily use. Grounded in the operation, not bolted on.

In practice

Workflow automation, custom agents, and decision tools — plus the data and process work that makes them reliable.

﹡How we work

Strong operations first, AI on top. Then it's yours.

AI creates real value when it's built on an operation that works. So we do both: we get the foundation into shape, then build the AI layer on top of it, so the business sees real results and the AI has something solid to stick to.

01

Diagnose

Understand how the operation actually runs, and find where AI can create real value.

02

Strengthen the foundation

Get the underlying process into shape.

03

Build the AI

Design and ship applied AI tools and agents in-house. Fit to your operation.

04

Transfer & govern

Train your team and set up governance so the result keeps working after we're gone.

Full-width image band — real operations photography with a dark / oxblood treatment
Where AI meets the operation it's supposed to improve.
﹡Where to start

One clear first step.

Most engagements begin the same way — a focused, fixed-scope diagnostic that's genuinely about AI, and that surfaces the operational groundwork underneath. You leave with a roadmap whether or not we build the next phase together.

Fixed scope, fixed price, a few weeks
Where AI creates value in your operation — and where it doesn't
The process and data work required to make it pay off
A prioritized roadmap you own, build-ready
The front door
AI Deployment Diagnostic

A structured first engagement that tells you what's real, what's ready, and what to build first — grounded in how your operation actually runs.

Book a diagnostic
﹡Selected work

What this looks like in practice.

A snapshot of where the approach is live today. We're adding named, attributable engagements as they mature.

Image · lighthouse
AI Deployment

From paper and spreadsheets to live AI workflows

Eighteen months moving a construction-management firm off paper and Excel onto a modern operating platform. Now their trusted advisor, we're integrating AI agents into live workflows — the clearest end-to-end picture of the ATP model.

Image
Operational Excellence

S&OP & planning cadence  food manufacturing

A structured S&OP process, demand-planning workflows, and inventory policy — the foundation an AI layer now compounds instead of automating chaos.

Image
Supply Chain

S&OP & planning cadence  food manufacturing

A structured S&OP process, demand-planning workflows, and inventory policy — the foundation an AI layer now compounds instead of automating chaos.

Note for us: prior OpEx/SC examples are optional and should be swapped for attributable ATP references before launch. Name CCM only with permission.

﹡About ATP

A boutique firm of hands-on operators.

ATP is a small firm by design. Between us we've spent careers inside operations and supply chains - advising leaders and, just as often, rolling up our sleeves to make the work actually happen. We founded ATP on a simple belief: the hardest part of AI in operations isn't the model. It's the operation.

We run ATP itself lean on AI rather than on a bench of analysts. We don't just talk about AI, we operate on it.

Operators who build

We don't just advise on technology — we've deployed and customized the systems businesses run on, led digital transformations, and built software ourselves. So the AI we bring, we can actually build and make work.

Independent

We're not tied to any one platform or model. We build with what fits your business, and we're straight about what will move the needle and what won't.

Built to be handed back

We transfer the capability and set up the governance so the work keeps running after we're gone. You own it — that's the point.

ATP is more than our initials.

Yes, it is who we are: Arbov, Tomashof, and Pelleg.

It is also where we help companies go: Accelerating to Performance.

It is how we help them modernize: AI Transformation Partners.

And it is what strong supply chains need to deliver: Available to Promise.

﹡The partners

Senior-led, every engagement.

BA
Ben Arbov
Partner

Has spent his career advising operations, supply-chain, and business leaders alongside Eli and Rafi — and building the software and tools that turn the advice into something real.

EP
Eli Pelleg
Partner

Thirty-five years at Tefen, one of the original operational-excellence consultancies — floor-level improvement across complex manufacturing and supply-chain operations.

RT
Rafi Tomashof
Partner

Former CEO of Programa consulting — decades leading operations and supply-chain transformation for complex businesses.

﹡Get in touch

Let's talk about your next steps.

Our partners are waiting to hear from you. Connect with us for a conversation on how you can use ATP to push your operation into the next phase.

*
Our Global Presence
New York - Israel - Europe
Send message
Thank you! Your submission has been received!
Oops! Something went wrong while submitting the form.