All glossary terms
B Operating model Economics

Build vs buy for agent platforms

Building an agent is not hard, and anybody telling you otherwise is selling something. The expensive part is everything around it, which is why this is two decisions rather than one, and why most organisations should answer them differently.

Definition

Build versus buy for agent platforms is the decision about who supplies the layer that runs, governs and records your agents. It is usually framed as one choice and is better treated as two separable ones: whether to build the agents themselves, and whether to build the control plane around them.

Why is ‘Build versus buy’ for agent platforms two decisions?

An AI agent and its control systems follow opposite rules. Agents are cheap to build and tailored to your business. They change constantly. The control layer is the reverse. It is costly to build and works the same everywhere. It must always keep running while everything above it is rebuilt.

Treating them as a single procurement question produces the two failures you see most often: an organisation that bought a platform to run three agents, and an organisation that built a control plane because its first agent was easy.

Two layers shown separately. The upper layer holds the agents: prompts, domain logic, tool choices and evaluation criteria, described as specific to your business, cheap to build and changing constantly, with the recommendation to build. The lower layer holds the control plane: registry, identity and scoping, audit trail, cost attribution, promotion gates, kill switch, observability and policy enforcement, described as identical across companies, expensive to build and required to keep working while agents change, with the recommendation to buy unless one of the listed exceptions applies.

The reframe matters commercially as well as architecturally. A vendor selling you the ability to create agents is selling you the cheap half, and a build proposal that scopes only the agents has estimated the cheap half. Both errors come from collapsing two decisions into one.

What should you almost always build?

The agents. They encode how your business actually works, which is the part no vendor can supply and the part you would not want them to. This is also, honestly, the easy half, and it has got easier every quarter since agent frameworks became commodity open source.

What sits in this layer is genuinely yours and genuinely straightforward to produce.

  • The prompts and the domain logic. What good looks like in your business, which nobody outside it knows.
  • Which tools each agent may call, and in what order. A decision about your systems and your risk appetite.
  • The evaluation criteria. What counts as a correct outcome here, which is the hardest thing on this list and still yours. See agent evaluation.
  • The workflow shape. Which steps need discretion and which should stay deterministic, covered at agentic workflow.

Worth stating plainly because it cuts against most vendor material. The agent framework layer is commoditised, open source and improving fast. A competent engineer can have a useful agent running in an afternoon, and the honest read of that is not that agents are trivial but that the difficulty was never there. If a proposal justifies platform spend by making agent creation sound hard, that is the claim to press on. The framework question is covered at agent frameworks.

What is expensive to build and rarely differentiating?

Everything that makes a population of agents operable rather than a demonstration. None of it is visible in a pilot, all of it is required in production, and no part of it distinguishes you from a competitor who has also built it.

Eight components, and the useful test is whether you would ever describe any of them to a customer as a reason to choose you.

Control plane components, what each costs to build, and whether it differentiates
Component What it takes to build well
Registry and discovery Auto-discovery across every platform agents get created in, not a form people fill in. This is the one most underestimated
Identity and scoping Per-agent credentials, intersection of agent and user permissions, and rotation. Hard to get right, catastrophic to get wrong
Audit trail Every model and tool call with actor attribution, unsampled, tamper-evident, retained for the period your regulator expects
Cost attribution Per agent, per model, per user. Aggregate spend is easy; the breakdown that lets anybody act is not
Promotion gates Dev, QA and production with version lineage, so a change is reviewable rather than a redeploy
Kill switch Per agent, cascading through everything downstream, operable by somebody who is not an engineer, at a weekend
Observability Step-level traces where a wrong answer returns a success code, which conventional monitoring does not catch
Policy enforcement Rules that hold at runtime rather than sitting in a document, which is the argument at agent governance

Two things about this list are worth noticing. Every item is invisible until you have more agents than you can hold in your head, which is why the build estimate made during the pilot is almost always low. And not one of them is something a customer would ever choose you for.

When does building the platform genuinely win?

Building your own control layer is the right choice more often than vendors admit. Five situations make the case for building. If two or more describe your business, the default flips. You then have to justify buying instead of building.

These are not edge cases. Several are common.

  1. 01
    The agent behaviour is the productIf what you sell is the agentic capability itself, the control plane is part of your product rather than overhead. Buying here means outsourcing your differentiation and inheriting somebody else's roadmap for it.
  2. 02
    You will never have many agentsPlatform economics assume a population. Three agents that rarely change do not amortise a platform, and the governance problem those three create is small enough to solve with a spreadsheet and a good habit.
  3. 03
    You already own the adjacent capabilityMature internal identity, a real observability stack, an existing service catalogue with owners. The marginal build is then far smaller than the greenfield estimate, because most of the control plane is integration rather than invention.
  4. 04
    Your constraints exclude the marketGenuinely air-gapped, a sovereignty requirement no vendor meets, or a stack nothing integrates with. Worth testing rather than assuming, and when true it settles the question.
  5. 05
    The abstractions are still movingBuying commits you to one vendor's model of what an agent is, during a period when that is being renegotiated. A thin layer you control can be cheaper to change than a platform you are embedded in.

The fifth point has become weaker over the past year and is worth revisiting rather than repeating. Open protocols have made agents more portable than they were, which cuts against the lock-in argument on both sides: it makes buying less committing, and it also removes some of the reason to build. See MCP and A2A.

What are you comparing?

Usually two things that are not comparable. A build estimate prices getting to working. A platform price covers staying working. Setting one against the other is the most common error in this decision, and it is the reason build cases look favourable at the moment they are approved.

The gap shows up in year two, and by then the decision is not revisitable at reasonable cost.

Two cost profiles over three years. The build line starts with a visible engineering spend in year one to reach working, then continues as ongoing but less visible cost: maintenance, on-call, keeping pace with model and protocol changes, and the opportunity cost of the team not working on the domain. The buy line is flat and visible across all three years. A note states that the build estimate prices getting to working while the platform price covers staying working, and that comparing them directly is the common error.
  • The build estimate is made by the people who would do the building. Not dishonestly. Engineers estimate construction well and maintenance badly, because maintenance lands on a future team with a different budget.
  • The ongoing cost is real and unbudgeted. Model providers change interfaces, protocols version, a compliance requirement arrives. Someone maintains that indefinitely, and it is rarely in the business case.
  • Opportunity cost is the largest line and never appears. A platform team building an audit trail is a platform team not building anything your customers notice.
  • Buying has an unbudgeted cost too. Integration, migrating what exists, and the work of changing how teams operate. A platform price is not a total cost either, and treating it as one is the mirror-image error.

How do you decide?

Forecast the population, then price the control plane honestly against it. The number of agents you expect in two years decides this more than any other input, because everything expensive on the control plane list scales with population rather than with sophistication.

Five steps, and the first usually settles it.

  1. Forecast agents in twenty-four months, not use cases today

    Under ten, across one or two teams, changing rarely: build, and keep it thin. Dozens, across many teams, created by people you do not manage: the governance problem dominates and buying usually wins.

  2. Cost the eight components, not the agent

    Take the control plane list and have somebody estimate each properly, including the second year. Compare that total against the platform price, because that is the like-for-like comparison.

  3. Ask who maintains it in year three

    By name, with the budget line. If the answer is the team that built it, check whether that team expects to still be doing this, and whether anybody has told them.

  4. Check the five build-wins conditions against yourself

    Honestly, not aspirationally. Two or more and building is probably right. None and you are proposing to construct something that will never distinguish you.

  5. Test the exit in both directions

    How would you leave the platform, and how would you migrate off your own build if it stalls. If neither answer exists, the decision is more committing than it looks and deserves more scrutiny than it is getting.

One pattern worth considering, because it is what several organisations settle on and it is rarely presented as an option. Buy the control plane and build the agents on top of it, which follows directly from the two-decisions framing. It keeps the part that encodes your business inside your team and hands over the part that is identical everywhere, and it is generally the cheapest position to change your mind from later.

Frequently asked questions about build vs buy

If your question is not here, our team will answer it directly.


Talk to a Specialist →
Should you build or buy an agent platform?

It is two decisions with opposite answers for most organisations. Agents are cheap to build, specific to your business and constantly changing, so build them. The control plane is expensive to build, identical across every company that needs one, and has to keep working while agents change underneath it, so buying usually wins unless one of a handful of conditions applies to you.

Is building an AI agent difficult?

No, and anyone telling you otherwise is selling something. The framework layer is commoditised, open source and improving quickly, and a competent engineer can have a useful agent running in an afternoon. The honest read is not that agents are trivial but that the difficulty was never there. If a proposal justifies platform spend by making agent creation sound hard, press on that claim.

What is the control plane in an agent platform?

Everything that makes a population of agents operable rather than a demonstration: registry and discovery, identity and scoping, audit trail, cost attribution, promotion gates, kill switch, observability and policy enforcement. None of it is visible in a pilot, all of it is required in production, and not one component is something a customer would ever choose you for.

What should you never outsource?

The agents themselves. They encode how your business actually works: the prompts and domain logic, which tools each may call and in what order, what counts as a correct outcome, and which steps need discretion rather than staying deterministic. No vendor can supply that, and you would not want them to, because it is the only part of the stack that reflects your business rather than the category.

When does building the platform actually win?

Five situations, and they are not edge cases. When the agent behaviour is the product you sell. When you will never have many agents. When you already own mature identity and observability, so the marginal build is integration rather than invention. When your constraints genuinely exclude the market. And when you judge the abstractions still to be moving. Two or more and building is probably right.

Why do build estimates come in low?

Because they are made during the pilot, when every expensive component is still invisible. Registry, identity scoping, unsampled audit, per-agent cost attribution and a cascading kill switch only become necessary once you have more agents than you can hold in your head. The estimate is also produced by the people who would do the building, who price construction well and maintenance badly.

What is the most common mistake in this decision?

Comparing a build estimate with a platform price. They are not comparable: a build estimate prices getting to working, while a platform price covers staying working. That mismatch is why build cases look favourable at the moment they are approved and the gap appears in year two, by which point the decision is no longer revisitable at reasonable cost.

What costs are missing from a build business case?

Maintenance as providers change interfaces and protocols version, on-call for something now sitting in a production path, and the compliance work that arrives on somebody else's timetable. The largest line is opportunity cost and it never appears: a platform team building an audit trail is a platform team not building anything your customers notice.

What costs are missing from a buy business case?

Integration, migrating whatever already exists, and the work of changing how teams operate, which is usually the largest of the three. A platform price is not a total cost either, and treating it as one is the mirror image of the error people make with build estimates. Both cases are routinely presented with only the visible half of the number.

Does buying a platform lock you in?

Less than it did, and this cuts both ways. Open protocols have made agents more portable than they were, which makes buying less committing and also removes some of the reason to build your own thin layer. The useful test is to write down how you would leave, in both directions. If neither answer exists, the decision is more committing than it looks.

How many agents justify buying a platform?

Forecast twenty-four months out rather than counting today's use cases. Under ten agents across one or two teams, changing rarely, and building thin is reasonable. Dozens across many teams, created by people you do not manage, and the governance problem dominates everything else. Population drives this more than sophistication, because every expensive control plane component scales with the number of agents.

Is there a middle option?

Yes, and it is what several organisations settle on while rarely being presented as a choice: buy the control plane and build the agents on top of it. It follows directly from treating this as two decisions. The part encoding your business stays with your team, the part that is identical everywhere is handed over, and it is generally the cheapest position to change your mind from later.

Price the eight, not the agent
Cost the control plane properly, then come and compare

SERAA Cortex supplies the eight components above: auto-discovery into one registry with version lineage, prompts, keys and rules scoped per tenant, promotion through Dev, QA and Production, an instant kill switch for any agent on any cloud, and every model and tool call recorded with the agent, the model, the user and the cost. Build your agents on top of it.