Relationships
Commerce is not transactions. Commerce is relationships.
Look at any company—yours, mine, the largest on earth—and strip away the org chart, the product, the brand. What remains is a web of promises between parties: who you do business with, what you agreed to, how the money flows, and ultimately who you trust.
The global economy may appear to be a ledger of transactions, but it is not. It is a graph of relationships, with hundreds of trillions of dollars moving across it every year.
That graph had no memory.
Relationship management in name only
CRM was the great half-victory in the attempt to digitize commercial relationships.
It’s one of the most valuable software categories—proof that remembering your relationships is worth a fortune. The concept of a relationship was present in the name only.
Open any CRM and look at what the system is organized around. The relationship is not the primary object. The atomic unit that powers every forecast, workflow, and commission plan ultimately maps to an “Opportunity”.
That nomenclature is a confession.
An opportunity is something you extract value from. A relationship is something you build and maintain. One is a trade, and the other an investment. CRM chose its worldview and wrote it into the schema: optimize the chase, ring the gong, and treat your counterparty as a row in a database; not a participant with a history and memory of its own.
So we built a global economy where each side keeps a partial, fragmented, and self-serving memory of the same relationship. CRM is transactional. Partial data, and care after close. The performance, billing deviations, and the signals that build trust to determine whether a relationship compounds or decays—are lost. Siloed in PDFs nobody reads, systems that can’t talk to each other, inaccessible inboxes, and the heads of people who eventually leave.
The most valuable asset in commerce is the relationship itself. Ironically, CRMs run on relational databases; they just never modeled the relationship.
Early…then right on time
This idea has been tried. I know because I was there. I was an early hire at RelateIQ, the company that coined the term Relationship Intelligence in 2013.
The conviction was right: the relationship, not the opportunity, is the real asset. This was in the pre-transformer world. Computers couldn’t yet satiate the ambitions of our intelligence appetite. RelateIQ still had a fantastic exit. Salesforce itself bought the company.
A relationship lives and grows in language. Agreements, emails, negotiations, exceptions, and commitments––this is where relationships are incubated, form, build, and endure or fade.
Relationships are a reading problem, dressed up as a data problem. And in the last three years, machines have learned to read and reason. Ideas don’t die; they wait. And as the idea waited, the market appetite has only grown.
The missing capabilities have arrived.
We are done waiting.
The wedge and the world
We started where the asymmetry was most brutal: buying.
Vendors have CRMs. Buyers have spreadsheets. We arm the unarmed side. "Power to the buyers" has been our rally cry.
But the categories were always wrong., Every company is a buyer, a seller, and a partner—sometimes with the same counterparty at the same time. Deel is our customer, our vendor, and a channel partner. Which row of which CRM does that relationship live in?
It does not. Relationships do not fit neatly in rows.
That is why the foundation of BRM is not a workflow tool that humans operate. It is an ontology built and maintained by agents: Parties, Agreements, and Commercial Models—the who, the what, and the how of every commercial relationship —structured so humans and AI agents can read it, reason over it, and act on it.
Commercial relationships exist between institutions, but they are carried by people. The champion, the negotiator, the executive sponsor, and the person who actually knows how the agreement came together all hold pieces of the truth. When those people leave, the relationship walks out the door with them unless something remembers it.
We call that something the Commercial Relationship Graph.
Vendor intelligence is the first thing we built on it. It will not be the last.
The graph starts with identity
A relationship graph is built upon identity.
The problem; there is no globally reconciled company identity. The same company appears differently across contracts, invoices, card transactions, identity systems, subsidiaries, marketplaces, resellers, billing entities, and geographies––and that’s just one company. Before an agent can understand a relationship, it has to know that all those fragments belong to the same party.
Today, BRM resolves vendor data from every system to a global BRM Company ID. That unique identity is built, owned and updated by us. own ontology, taxonomy, and proprietary dataset. We are a data company, masquerading as an application company. We do not rent our foundation from a third-party data provider.
If data is the fuel for AI, we must get our data foundation correct first; it’s table stakes. It’s also difficult.
Our Listings Engine continuously discovers, classifies, and enriches companies with BRM native categories, descriptions, products, and commercial context. The world changes too quickly—and the underlying data is too fragmented for armies of people to curate by hand. So we built an army of Agents. This is a machine-scale intelligence problem.
That foundation is why BRM can serve global enterprises with billions in spend.
The global layer establishes who a company is—and it is shared across every BRM. Each customer's instance remembers what that company means to them: their agreements, communication, spend, history, decisions, and accumulated context. Company-specific data remains inside that company's BRM––it is not shared across the network.
The identity is global. The relationship remains yours.
Relationship Intelligence, realized
Here is what becomes possible when the graph exists.
A relationship becomes a living, breathing, up-to-date memory.
An agent reads every agreement you have signed with a counterparty, every exception you approved, and every commitment either side kept or broke. It knows internal and external sentiment. It understands the history of that relationship as if that were its only job—because it is. Renewals negotiate from history instead of amnesia. Risk surfaces before detonating. Trust becomes grounded in evidence, not institutional folklore. And informed action happens on time, every time. The agents that operate the graph don’t tire.
The graph also reaches the most important layer: the human layer. Your company already produces the signals: the meeting where a commitment was made, the email that changed the terms, the reply that came immediately, and the one that never came at all. Today, most of those signals evaporate. On the graph, they become legible. You can see which relationships are strengthening, where momentum has disappeared, when a champion has quietly disengaged, and who actually knows whom when a door needs to be opened.
Relationship strength has always hidden in the exhaust of ordinary work. For the first time, machines can read it, reason over it, and help you build them.
That is Relationship Intelligence: the accumulated truth of a relationship, made permanent, legible, and actionable.
Sellers got a partial version of this in 1999 and built trillions of combined market cap. The whole thing—across roles, across systems, across the full history of the relationship, built natively for agents—has never existed.
We believe it is one of the great unbuilt layers of the economy.
The internet gave commerce a network. Payments gave it rails. Nobody has given it a memory.
The agent economy has an amnesia problem
Agents are negotiating, corresponding, and transact in our names.
But an agent that does not know your history, your commitments, and your judgment is not representing you.
It is impersonating you.
The rote work of commerce—the chasing, the reconciling, the redlining of basic terms at midnight the night before—will fade into the background. Agents will do it tirelessly and well. When execution becomes abundant, the scarce thing becomes judgment: who you decide to do business with and how you keep faith with them.
Do I trust this counterparty? Have they kept their word? Is this a relationship worth building for a decade? The questions you never had time to answer were = the questions that matter most.
As agents begin talking to agents, the digital paper trail will not shrink. It will explode—far beyond what any human can hold in their head. There will be more communication, more agreements, more decisions, and more signal than ever. Somehow, it all has to remain true to who you are, what you agreed to, and how you conduct business.
That requires more than a filing cabinet. It requires a living, structured, machine-legible record of every relationship, and every commitment inside it.
The Commercial Relationship Graph is how people remain in command of an agent-speed economy. Agents can execute, but they need a memory. They can recommend and negotiate, but people must remain accountable for the judgment that defines who they are willing to trust.
The agents act. The graph remembers. The human decides.
Hard problems worth solving
Building this is not a matter of putting a chat box on top of a contract repository, nor an LLM in front of an ERP.
The graph has to resolve fragmented identities across thousands of systems. It has to reconcile contradictory sources, distinguish promises from performance, preserve provenance, understand how relationships change over time, and know when something is true, inferred, outdated, or simply unknown.
It must give agents enough context to act without allowing them to invent institutional history. It must understand permissions, because remembering everything does not mean showing everything to everyone. And it has to be right when the source material is messy, the stakes are high, and the answer is inconvenient.
These are not solved problems.
That is precisely why they are worth solving.
The ones who build it
Ten people. Brannan Street. Building the memory layer of commerce.
Say that out loud and it sounds absurd. It's supposed to. Every company I've ever admired sounded absurd right up until it sounded obvious, then everyone claims they saw it coming. That gap in between is where the work actually happens.
It takes a specific kind of person to stay in it.
Tenacity—because a twenty-five-year head start does not yield politely.
Agency—because at ten people, no one waits to be told.
Progress—because we demo what is real every week, not what is planned.
Excellence—because global companies are betting their commercial memory on us being right.
If you have ever looked at CRM—the defining enterprise software category of the last generation—and thought that is only half the product, you already see what we see.
The other half is unclaimed.
I have waited more than a decade for the tools to build it. The tools are finally here. The foundation exists. The team is growing.
There is a place for you.
Power to the buyers was the battle cry.
Giving commerce a memory is the mission.
- Cuddy, Founder & Chief Relationship Officer
