The board asks: build or buy? The CIO answers with a table, two columns, twelve rows, a total at the bottom. Both columns end the same way. In three years, the intelligence is rented by usage, the infrastructure is either yours to maintain or theirs to reprice, and the model of your business sits in a format someone else controls.
Ask what you will own
Build or buy is a procurement question, with a procurement answer: whichever is cheaper this budget cycle. The question that matters over ten years is different: what do you own at the end?
Models converge quickly. Compute is a commodity. Talent moves. What does not move is the explicit model of your own operations: your entities, your rules, the exceptions added after past incidents, the vocabulary your teams actually use.[1] It is the last lasting advantage, and it grows in value: every exception resolved is added to it.
So ask each option the same question: where will the model of my business live, who can read it, and who can take it away?
Build: own the infrastructure, rent the intelligence
The internal team builds the orchestration, the connectors, the guardrails. After two years you own a lot of infrastructure. The intelligence inside it is rented, per use, from a supplier who can change the model, the price, the rate limits or the terms overnight.[2] An in-house pipeline pays that inference bill exactly like a purchased one.[3]
Open-weight models do not change this. Weights trained elsewhere, on data chosen elsewhere, according to priorities set elsewhere, are a better rental agreement, nothing more.[4]
The larger cost is what happens to the model of your business. The business writes tickets, the platform team turns them into code, and meaning is lost at each step without anyone noticing.[5] After two years, your operating model exists as code that the people who do the work cannot read, written by people who are moving on.
Buy a US platform: rent both
You rent the intelligence and the infrastructure, and you configure the model of your business inside someone else’s system. Enterprise software has always required the organisation to adapt to the product. Over time, the platform shapes the company more than the company shapes the platform.[6]
The configuration is done by an integrator billing by the day, whose business depends on the gap persisting.[7] The commercial model is public: pilots and bootcamps at little or no cost to the customer, converted into long-term contracts and expanded account by account. It is a sound business, and the asset that grows in value accrues to the platform.
At the end, your ontology exists, in the vendor’s format, under the vendor’s licence, readable by the vendor’s tools.
The third option
Own the model of your business. Run it inside your own infrastructure. Pay for the value it produces.
With Monarch, the ontology belongs to you and you can inspect it, written in the words of your teams. It runs under your accounts, your keys and your jurisdiction; Galahad hosts none of your data, and Monarch can run fully isolated on a model you host yourself.[8] You are invoiced after the result: a share of the value created, payable once in production. Hours, tokens and seats are all billed up front, and none of them commits the vendor to a result.[9]
Galahad refuses three things. To host your data, because an audit trail you have to request is not yours. To hold your model in a format you cannot read, because the asset must grow for you. To bill before production, because a vendor paid up front has already been paid for your pilot.
What each option costs
Building costs a permanent team, plus inference. Buying costs a licence and an integrator, plus inference. Both are paid whether the result is there or not.
The third option costs a share of the value produced, and nothing if there is none. It aligns the vendor’s interest with yours.
Questions to ask a vendor
Ask them in writing, before the demonstration.
- When the contract ends, in what format does the model of my business leave, and who can read it?
- Where does inference run, and who can reprice it, throttle it or cut it off?
- Who writes the rules: my teams, or your integrator?
- Can this run fully isolated, on my network, on a model I host?
- What do I pay before anything is in production?
- What happens to what you learn inside my organisation?
A vendor who answers all six leaves you owning something.
Monarch is the third option: the model of your business as an asset you own, running inside your infrastructure, paid for on the value it produces. See it on your data.