When a delivery truck arrives at a construction site near Riyadh with bulk cement, the driver photographs the delivery note and sends it via WhatsApp. For most of BRKZ's existence, this triggered a chain of manual tasks: someone read the document, matched it to an order, verified quantities, updated systems and closed the delivery. Today, an agent called Nusa handles most of that work—reading the image, extracting details, matching them against internal systems, verifying amounts and completing the reconciliation. The team now focuses only on uncertain cases and exceptions. Each month, Nusa closes a growing share of deliveries.
Bulk cement may not seem like a glamorous AI application, but that's precisely the point. The real opportunity lies not in headline-grabbing use cases but in industries where technology has historically lagged: physical goods, fragmented supply chains, credit and logistics. The conversation among founders and investors has shifted from chatbots and headcount reduction toward a more fundamental question: can AI reshape how a company actually operates?
Foundations before intelligence
When BRKZ launched, the building-materials sector faced a more basic problem than needing AI. The industry required systems at all. Procurement was deeply manual and scattered across phone calls, WhatsApp, spreadsheets, PDFs and individual knowledge. Requests for quotation, pricing, supplier relationships, orders, deliveries, credit and collections existed nowhere in a unified form.
The company's first three years focused on less fashionable but essential work: moving core workflows onto shared platforms. How demand entered the system, how suppliers were assessed, how quotations were generated, how orders were processed, how deliveries were verified, how customers paid and how every transaction was recorded—all of this moved into digital systems.
This created something valuable: every workflow began generating a digital record. Each RFQ, quotation, supplier interaction, delivery and payment became data. Over time, these records accumulated into tens of millions of structured data points covering products, suppliers, transactions and payment patterns. The company didn't require contractors and transporters to change their communication methods. An RFQ could still arrive on WhatsApp and a delivery note as a photograph. Agents transformed those inputs into structured records the business could act upon.
The fundamental principle emerged: "You cannot add intelligence to a business you have not made observable."
The progression: systemize, capture, understand, automate, agentize
The AI journey for traditional industries follows a clear sequence. First comes systemization—establishing shared platforms. Then data capture—recording every workflow. Understanding follows—analyzing patterns in that data. Automation comes next—handling routine tasks. Finally, agentization—giving agents responsibility to act within defined boundaries.
The first two steps are non-negotiable. Without data, AI lacks proprietary context. Without context, a general-purpose model sits atop information available to everyone else, offering no competitive edge.
From answering questions to taking action
Once BRKZ accumulated sufficient transaction history, new questions became possible. Could the system learn the company's pricing logic? Could it identify which suppliers matched a given RFQ?
Pricing building materials proves surprisingly complex. The same product carries different prices based on quantity, location, delivery needs, timing and market conditions. Traditionally, experienced procurement staff held this knowledge in their heads.
The company built a pricing engine using historical RFQs and transactions. Mizan, the AI procurement agent, now operates across initial product categories. It generates price recommendations in seconds for procurement officers to review and adjust before sending. Years of collected data shifted from describing what had happened to helping decide what should happen next.
Prediction has value. Execution matters more. Nusa demonstrated that AI becomes far more powerful when it can act within a system rather than simply answer questions about it.
Rethinking where agents fit
Instead of asking where AI could be added, BRKZ started asking which parts of transactions agents could own. A transaction naturally divides into domains—sales, procurement, credit, operations and finance—each with inputs, decisions, actions and exceptions. In procurement, an agent can interpret RFQs, identify products, forecast prices, select suppliers and generate competitive quotes. In credit, it can support underwriting, limits, exposure monitoring and collections, with human oversight for high-stakes decisions.
In operations, agents coordinate orders, validate deliveries and reconcile the physical world with digital systems. The goal is an agentic layer spanning the entire transaction, with each agent carrying clear responsibility for its work.
The invisible cost of coordination
Companies bear more than labor costs. Coordination costs accumulate silently. A single customer request might pass through sales, procurement, operations, logistics, finance and credit before completion. Each handoff introduces delay. Someone sends a message. Someone requests approval. Someone copies information between systems. As organizations scale, coordination itself becomes work.
AI can eliminate vast amounts of this invisible work, and that impact will likely exceed the value of automating individual tasks.
Amplifying human capability
The most important insight from conversations with US founders wasn't "how many people can AI replace?" but rather "how much more can every person accomplish?" The goal is turning a skilled salesperson into someone with capabilities that would previously have required an entire support team.
Consider a salesperson with a digital twin. It knows which customers are likely to reorder, which have quietly reduced purchases, which quotations didn't convert and which products a customer should be buying but isn't. Who should I call? Why now? What should I sell, and at what price? Which relationship is deteriorating before I've noticed?
The salesperson retains ownership of the relationship. An agent handles preparation, analysis and routine coordination, and eventually manages follow-ups and reorders within agreed limits. The same logic applies across finance, procurement and operations. As agents take on routine work, people focus on where judgment and relationships create the most value.
Making the physical world visible
In physical industries, much of what matters still happens outside a company's captured data. A truck arriving at a site is data. A pallet being unloaded is data. A material failing quality tests, inventory sitting in a warehouse, a supplier repeatedly arriving late—all data. But if nobody captures these events, they don't exist digitally. AI can only reason about what it can see.
This suggests a counterintuitive idea: operational complexity can become a moat. If one operations employee can oversee dramatically more transactions because agents handle routine coordination, and one procurement person can manage dramatically more spend, the economics shift. The harder a workflow is to execute manually, the more valuable it becomes once machines can execute meaningful parts of it reliably.
Building what matters
Not every problem requires the most advanced model available. The enterprise AI stack won't contain one model but many, with architecture determining which—local, specialized or frontier—fits each task. AI makes building software easier, but if someone has already solved a generic problem extremely well, purchasing beats building. BRKZ directs engineering capacity toward what the company owns: workflows, transaction data, pricing knowledge, supplier networks and payment history.
Competitive advantage stems less from which model is used and more from what context can be provided and what actions agents are permitted to take. AI doesn't diminish strategy. It makes deciding what not to build more important.
The connected transaction
The larger opportunity lies in connecting the entire transaction across sales, procurement, credit, operations and finance. A customer shouldn't need to understand BRKZ's organizational structure. They should say, "I need these materials at this project next Tuesday," and everything behind that request should be coordinated across agents.
Extend this five years forward. A contractor's procurement agent knows what the project requires and communicates with BRKZ's agent. BRKZ's agent understands the requirement, the customer's history, pricing and supplier availability, and talks to supplier agents. A logistics agent coordinates delivery. Financial systems agree terms. Humans set policy, manage relationships and handle exceptions. Machines execute the transaction.
Agents will transact with agents. Companies building infrastructure for that world today will look fundamentally different from companies adding chatbots to existing software. The technology isn't sitting beside the business model. It's becoming the operating system of the business model.
Measuring what matters
Over the coming years, every company will claim to be AI-powered, making the term nearly meaningless—like saying a company uses the internet. The real questions are different. Can the company process significantly more transactions without growing headcount proportionally? Can each salesperson generate more output? Can quoting accelerate, credit decisions improve, working capital turn faster and customers transact with less friction?
These are the AI metrics that matter, not tokens consumed or copilots deployed. The question that keeps returning: how much more output can BRKZ generate from every colleague, every dollar of working capital and every transaction because of this technology?
An opportunity for the region
This matters particularly for founders in Saudi Arabia and the broader MENA region. Enormous industries remain early in digitization: construction, manufacturing, logistics, healthcare and wholesale trade. Many view this as a technology disadvantage. The opposite is true.
There's a chance to skip an entire generation of software. The region doesn't need to spend twenty years reproducing every system and workflow built elsewhere. These industries can be systemized now, on the assumption that intelligence and agents will sit inside those systems from day one.
If an industry still runs on WhatsApp, Excel, phone calls and people's heads, start by making the workflow observable. Own the proprietary context of the business. Then ask where intelligence can improve decisions and, once reliable, where it can move from advising people to doing the work.
This is an extraordinary opportunity for the ecosystem. But it requires founders to share more of what actually works and what doesn't: the architecture, the mistakes, where humans still outperform machines, where agents fail, and where the economics do and don't work. The ecosystem will move faster if we learn from each other.
Starting from resources, not features
BRKZ remains early in this journey. The company is experimenting, learning and making plenty of mistakes. But one principle has become clear.
Don't start by asking how to add AI to a company. Start by asking how the company would be built differently if intelligent machines were simply another resource available alongside people, capital and software. Then work backwards.
For BRKZ, the path has been: systemize, capture, understand, automate, agentize. For another company it may look different. But the objective should be identical. Don't use AI to make the old way of working slightly more efficient. Use it to discover a fundamentally better way of operating.
If more founders in the region approach it that way and openly share what they learn, the impact will extend far beyond any one company. That's the conversation this perspective aims to contribute to.
Source: The Next Web
