AI Agents

Most AI agent work should not be bought from a consultancy. Here is the part that should.

Gartner expects over 40% of agentic AI projects to be canceled by 2027. We will tell you which product to buy when a product wins - and only build where one genuinely cannot reach.

AI agents are the busiest category in enterprise software right now, which is exactly why most of the money spent on them is wasted. The useful question is not "should we use AI agents" but "which parts of this are already a $99-per-month product, and which parts will no vendor ever build for us?" Getting that line wrong in either direction is expensive.

The failure numbers, and what actually causes them

In June 2025, based on a poll of more than 3,400 organizations, Gartner predicted that over 40% of agentic AI projects will be canceled by the end of 2027, citing escalating costs, unclear business value, and inadequate risk controls. The same firm expects agentic AI to make around 15% of day-to-day work decisions by 2028, up from essentially zero in 2024 - so the technology is not the problem.

The MIT report The GenAI Divide: State of AI in Business 2025 put it more bluntly: roughly 95% of pilots delivered no measurable profit-and-loss impact, drawn from 52 executive interviews, 153 leader surveys, and 300 public deployments. That figure got repeated everywhere, usually without its two most useful findings:

Underneath all three of Gartner's stated causes sits one root problem: the project began as a capability demonstration rather than as a named metric with a baseline. Nobody agreed in advance what success would look like, so nobody could prove it happened, so the budget got cut. That is the specific failure our pricing model is designed to make impossible - if we cannot define and measure the metric, we do not get paid, so we will not start.

Buy this. Do not hire us for it.

We would rather lose the project than sell you a build you can replace with a subscription. Where products already win:

The commoditization pressure here is real and it is aimed squarely at consultancies. A Google Cloud executive said publicly in February 2026 that thin-layer LLM wrapper companies face existential risk as foundation model providers absorb their functionality, and Gartner has been reported as expecting around 40% of enterprise applications to embed vertical agents by the end of 2026. Any firm still selling "we'll build you a chatbot" in that environment is selling you a depreciating asset.

Where custom agents still win

Products are built for the workflow that most customers share. The money a product cannot reach sits in the parts of your operation that are specific to you:

How we measure an agent, and why deflection is a bad metric

Most agent reporting optimizes for the wrong thing. Deflection rate rewards an agent for ending conversations, not for resolving them - a system that frustrates people into giving up scores beautifully on it. What we will actually agree to be paid against, in order of how verifiable they are:

MetricWho verifies itStrength
Recovered money (receivables collected, unbilled exceptions invoiced) The paying counterparty, then your AR Strongest - we cannot influence or estimate it
Fully resolved contacts with no human handoff Your ticketing system, against a pre-deployment baseline Good, if resolution is defined before launch and re-contact within a window counts as unresolved
Staff hours returned on a named process Your own time and volume records over matched periods Workable, and the number most often inflated - so we hold it to the same baseline discipline
Deflection rate, containment rate, messages handled Nobody meaningful We will not price on these, and neither should you

Guardrails we build in by default

The honest summary

AI agents are a genuinely enormous market and most of it is not ours to sell. If your need is a front desk, an SMB inbox, or standard ticket deflection, buy the product - we will name it on the call and charge you nothing. If your need is the multi-system, exception-ridden, legacy-bound operational work that no vendor will ever productize, that is a real project, and we will take it on with no upfront fee and get paid from what it recovers.

Sources

Every figure on this page is linked to its origin so you can check it. Vendor marketing claims are labelled as such in the text rather than presented as research.

  1. Gartner, Predicts Over 40% of Agentic AI Projects Will Be Canceled by End of 2027 (June 2025, poll of more than 3,400 organizations) - the cancellation prediction and its three stated causes.
  2. MIT NANDA, The GenAI Divide: State of AI in Business 2025 - reported via The Hill. Source of the ~95% no-measurable-P&L-impact finding, the vendor-versus-internal build comparison, and the sales and marketing budget concentration.
  3. Meta's WhatsApp Business AI launch for Brazilian SMEs, early 2026 - reported via Exame and MobileTime, including the Kantar figures cited by Meta.

Founding client terms

We are early, we want proof, and we are pricing accordingly

Book the free audit We run three performance-based engagements at a time

Frequently Asked Questions

Why do AI agent projects get canceled?

Gartner predicted in June 2025, from a poll of more than 3,400 organizations, that over 40% of agentic AI projects will be canceled by the end of 2027 - citing escalating costs, unclear business value, and inadequate risk controls. Behind all three sits the same root cause: the project started as a capability demonstration rather than a named business metric with a baseline, so nobody could prove it worked.

Should I build a custom AI agent or buy one?

Buy whenever a productized vendor already serves your exact workflow - front-desk answering, SMB messaging, and generic support deflection are commoditized, with products roughly $49-$300/month. Build when the agent must reconcile across systems no single vendor covers, when the value is in the exceptions products treat as out of scope, or when your core software is legacy enough that no connector exists.

Do Brazilian small businesses need a custom WhatsApp AI agent?

Usually not. Meta released WhatsApp Business AI to Brazilian SMEs in early 2026 after testing in Mexico. It answers customers 24/7 from your catalog, website, and stored policies with no programming, no integration, and no extra cost. Custom work only makes sense above that ceiling - where an agent must write to inventory, ERP, or booking systems Meta's tool cannot reach.

How do you measure the ROI of an AI agent?

Prefer metrics an outside party verifies. Recovered money is strongest - receivables collected or unbilled exceptions invoiced, because a third party pays it and it lands in your AR. Fully resolved contacts without human handoff and staff hours returned are workable against a pre-deployment baseline. Deflection rate is a bad metric: it rewards an agent for ending conversations rather than resolving them.

Do most enterprise AI projects deliver returns?

MIT's The GenAI Divide: State of AI in Business 2025 found roughly 95% of pilots delivered no measurable P&L impact, from 52 executive interviews, 153 leader surveys, and 300 public deployments. Two under-quoted findings matter more: tools built with external vendors succeeded about twice as often as internal builds, and budgets concentrated in sales and marketing, where measured ROI was lowest.

What guardrails should an operational AI agent have?

A confidence threshold that escalates to a human with full context instead of guessing; no authority to make financial or contractual commitments autonomously; a complete audit log of actions taken and data read; and read-only access until the value is demonstrated. Inadequate risk controls is one of Gartner's three named causes of cancellation, and good logging is also what makes results measurable enough to price on.

Not sure whether to buy or build?

Free audit. If an off-the-shelf product already solves it, we will name the product and charge you nothing.

Schedule the Free Audit

Other operations we work in