The operational harness is the same guides-and-sensors pattern applied to running an entire business, not just code. That's Capiva's read. And it's where the return is: MIT's study (NANDA, 2025) shows the biggest AI ROI sits in the back office, not in sales or marketing, and that most pilots never reach the P&L.
The visible part and the bigger part
Almost everyone talking about harness engineering today means code: Claude Code, Cursor, agents that read a repository and run tests. That's the visible part, and it's maturing fast. The bigger part is the operational harness: the same guides-and-sensors pattern applied to running a whole business, AI beyond code. Support, finance, internal knowledge, back office, the processes that keep the company running every day. That's Capiva's read, not a finding from a report. And almost no one is engineering that layer with any discipline.
The diagnosis: why AI doesn't reach the P&L
The numbers explain the gap. McKinsey (State of AI, November 2025) says no more than 10% of companies have scaled AI agents in any given function, and only 39% report enterprise-level EBIT impact despite deployed use cases. Gartner (June 25, 2025) projects that more than 40% of agentic AI projects will be canceled by the end of 2027, citing unclear value and weak risk controls, and calls part of the market "agent washing". Deloitte (2026) found only 21% have a mature agent-governance model. The pattern repeats: everyone runs a pilot, almost no one reaches production.
Coding harness vs operational harness
| Dimension | Coding harness | Operational harness |
|---|---|---|
| Maturity | Tools ready, adoption growing | Early, almost all still to build |
| Examples | Claude Code, Cursor, capivaOS | Support, finance, knowledge, back office |
| Where the ROI is | Productivity of people writing code | Operations automation (the biggest return, per MIT) |
| Who builds it | Engineering teams | No one yet, at most companies |
| Main gap | Agent reliability in the repo | Integration and memory in real operations |
Why it's engineering, not the model
MIT (NANDA, 2025, via Fortune) points at the root cause: pilots don't die from model quality, they die from integration, a "learning gap" between the pilot and real operations. The same study shows two uncomfortable facts. About 95% of generative AI pilots produce no P&L impact, and the biggest return is in the back office, while more than half of budgets chase sales and marketing. And buying tends to beat building from scratch: purchased solutions succeed around 67% of the time, versus roughly a third for internal builds. The bottleneck is the engineering around the model. It's the harness that's missing.
Where Capiva comes in
We solve the coding layer with capivaOS, the open-source harness on top of Claude Code. The operational layer we build in production, with the same pattern of guides, sensors, memory, and governance, and the account is in AI Harness in Practice. The practical entry point is a strategic diagnosis: map where operations have real return before automating anything.
Frequently asked questions
What is an operational harness?▼
It's the same harness engineering pattern (guides, sensors, memory, verification, and governance) applied to running an entire business, not just code. That's Capiva's read, not a definition from a report. Where the coding harness handles agents that write software, the operational one handles AI in support, finance, knowledge, and the back office.
Why do so many AI projects fail to return value?▼
Because they stop at the pilot. McKinsey (November 2025) says no more than 10% of companies have scaled agents in any function, and Gartner projects more than 40% of agentic AI projects canceled by 2027. The root cause, per MIT, is integration, not model quality.
Where is the biggest AI ROI?▼
In the back office, per MIT's study (NANDA, 2025). The counterintuitive part is that more than half of budgets go to sales and marketing, where the return is smaller. Operations automation (internal processes, data, support) is where the P&L moves.
Should you buy or build the AI solution?▼
MIT's study suggests buying tends to work better: purchased solutions succeed around 67% of the time, versus roughly a third for internal builds from scratch. That doesn't mean never build. It means the value is in the integration and the harness around it, not in reinventing the model.
What's the difference between a coding harness and an operational harness?▼
The coding one is the visible, more mature part: Claude Code, Cursor, capivaOS, agents that read a repo and run tests. The operational one applies the same pattern to the entire business, and it's almost all still to build. One handles people writing software, the other handles how the business runs.
Where do I start with an operational harness?▼
With a diagnosis: map where operations have real return before automating, and look at governance early (only 21% of companies have a mature model, per Deloitte, 2026). Then run the same loop as the coding harness: each error becomes a rule, guides direct, sensors verify. The coding layer already has capivaOS ready on top of Claude Code.
Anyone can buy the model. Almost no one has built the operational harness yet, and that's where the return is.