Automation Is Not Architecture: The Thermometer vs Kidney Test
- Sunil Dutt Jha
- Oct 14
- 3 min read
When the Fever Is in the Enterprise Body
Every quarter, a new automation wave sweeps through the industry — “AI-driven monitoring,” “autonomous bots,” “self-healing infrastructure.” Yet incidents keep rising, especially in complex enterprises like Banks, Telcos, Oil majors, and Hospitals.
The truth?
Automation doesn’t fail because it’s weak. It fails because it’s misplaced.
The Fever Is Not in the Thermometer
In medicine, you don’t treat a fever by replacing the thermometer.
You find the organ that’s failing — the kidney, liver, or heart that’s creating the imbalance.
Enterprises are the same.
Automation systems are thermometers — they measure and react.
The organs — Finance, Operations, Risk, Sales, and IT — create the underlying behavior.
When automation keeps breaking, it’s rarely the tool that’s weak. It’s the architecture that’s misaligned.
The enterprise kidney is failing, and everyone’s staring at the thermometer.
Banking Example: The Overnight Settlement Glitch
A regional bank automated its settlement process. A batch automation engine was configured to trigger reconciliation jobs after trading close.
One night, settlement mismatches delayed downstream postings by six hours, freezing $180 million in transactions. At 2 AM, $180 million in trades got stuck.
Dashboards showed “All systems healthy.” Bots kept running. No alerts fired.
Six hours later, treasury operations realized nothing had posted.
Teams escalated instantly:
Monitoring dashboards showed green.
The automation scripts were “functioning as expected.”
The AI alert system even flagged a “resolved anomaly.”
But none of them fixed the issue—because the problem wasn’t automation.
It was architecture.
When traced through the Enterprise Anatomy Model, this was the result:
Finance (D3) and Operations (D5) had separate cut-off rules.
Risk (D7) had not updated the timing logic in its rule engine.
IT (D11) automated around old assumptions.
Automation wasn’t wrong — it was obediently enforcing outdated structure.
Why CEOs Must See the P1–P6 Map
Dissecting the Breakdown Using P1 (Strategy)–P6 (Operations)
Perspective | What Should Exist | What Actually Existed | Resulting Effect |
P1 – Strategy | Clear business outcome: “All settlements finalized by 2 AM with <0.01 % mismatch.” | Fragmented ownership between Treasury (D3 Finance) and IT Ops (D5 Operations). | No unified outcome definition. |
P2 – Process | Cross-department flow with exception handling for delayed trades. | Automation job chained to time, not business event. | Late trades never entered reconciliation. |
P3 – Systems / Logic | Rule logic aligned with trade window extensions. | Legacy rule engine still used fixed-hour cutoff. | Correct data excluded from automation scope. |
P4 – Component Specs | Parameterized batch job contracts. | Hard-coded scripts. | Changes required code edits, not configuration. |
P5 – Implementation Tasks | Version-controlled changes mapped to logic and contracts. | “Quick fixes” pushed manually during audit window. | Traceability lost. |
P6 – Operations | Cause-effect dashboards linked to business KPIs. | Tool-level uptime dashboards only. | Fever seen, organ unseen. |
Diagnosis:
Automation was the thermometer—it showed “normal temperature.
”But Finance (D3) and Operations (D5), the enterprise kidneys, weren’t filtering correctly. The enterprise body was in silent failure.
CEO Insight
When CEOs frame such incidents as IT failures, they prescribe more thermometers. But, when they reframe them as anatomy failures, they start rebuilding organs.
Ask your team after the next automation glitch:
Which enterprise organ failed—Finance, Risk, or Operations?
Which P-level was absent—Strategy, Logic, or Component?
What will we fix structurally before we automate again?
That’s the difference between an enterprise that survives incidents and one that never repeats them.
Action Item:
Every alert, every outage, every fever is a diagnostic opportunity.
Don’t buy a smarter thermometer.
Discover the kidney and understand how it works with lungs and liver.
Automation is not architecture. The thermometer is not the kidney.
