RS
RILSEY & INTELLIGENCE 真智卓见
Case Studies

Results, retestable.

Not slideware. Each engagement below shows the ANOI-6 entry score, the exit score, what we built, and what moved. All names are anonymized — the numbers are not.

不是 PPT。每个案例都标注 ANOI-6 入口分、出口分、我们建了什么、什么真的变了。客户已脱敏，数字没有。

Disclosure All case data is disclosed under written client authorization. Industry, scale, and outcomes are factual; identifying details and the deployment name have been anonymized per the client's NDA scope.
Case 01

National retail chain

3,000+ stores · ¥5B+ annual revenue · China
Insight + Build 14 weeks Workflow + Agents
Situation

The client had deployed a demand-forecasting tool across 3,000 stores. Store managers were overriding AI recommendations 80% of the time. The CIO insisted the model was accurate; the COO insisted the managers knew better. Both were right — the workflow around the tool was broken.

ANOI-6 · Entry vs. Exit (composite 2.1 → 3.5)
DimensionEntryExit
Strategy2.13.5
Data2.43.7
Workflow1.33.3
People1.63.1
Governance2.03.4
Stack3.23.8
What we delivered
  • Store replenishment workflow redesign
  • Demand-forecasting agent into production
  • Decision-rights matrix for store managers
  • 90-day handover to client ops team
What moved
72h → 6hStore replenishment decision cycle, end to end.
−31%Stock-out incidents at top-quartile stores within one quarter.
100%System operable by client team without Rilsey present after 90 days.
"The diagnosis told us what we already suspected but couldn't prove — that our stack was fine and our workflow was the problem. Six months later the score moved because we fixed the workflow, not because we bought more tools." — VP Operations, anonymized
Case 02

Regional commercial bank

4,000 employees · Mid-tier · China
Insight + Roadmap 10 weeks Governance + Compliance
Situation

The bank had been running AI pilots for 18 months across four departments. Each pilot succeeded in isolation — none had crossed into production. The board wanted one question answered before approving the next budget: is this a tool problem, a people problem, or a governance problem?

ANOI-6 · Entry vs. Exit (composite 2.3 → 3.6)
Strategy2.33.6
Data2.13.4
Workflow2.43.5
People2.23.3
Governance1.93.8
Stack2.73.6
What we delivered
  • 18-month transformation roadmap
  • Compliance-review automation pilot (live)
  • Model-decision accountability framework
  • Vendor-neutral evaluation of 4 platforms
What moved
60%Compliance review cases auto-covered by the pilot within one quarter.
−44%Median review turnaround time for in-scope cases.
0Vendor relationships locked in — platform choice remains the client's.
"We came in expecting a tool recommendation. We left with a roadmap that let us choose the tool ourselves, with the reasoning on paper. That's worth more than any RFP answer." — Head of Risk, anonymized
Case 03

Precision manufacturer

1,200 employees · Advanced manufacturing · APAC
Insight only 6 weeks Diagnosis + Roadmap
Situation

The client had invested more in AI tooling than any peer in their segment — their Stack score was 3.8 at entry, the highest of any client we have diagnosed. Yet their AI initiatives had produced zero measurable margin improvement. They came to Rilsey to find out why.

ANOI-6 · Entry (composite 2.4) — exit retest at 6 months (3.3)
Strategy2.63.3
Data2.43.2
Workflow2.13.3
People1.42.9
Governance2.33.2
Stack3.83.9
What we delivered
  • ANOI-6 diagnosis + evidence working papers
  • Prioritized opportunity map (7 initiatives)
  • People & skills gap analysis by role
  • Stack portability audit (vendor-neutrality)
What moved
2.4 → 3.3Composite ANOI-6 score at 6-month retest, post-roadmap execution by client's own team.
People 1.4 → 2.9Largest single-dimension gain — the weakest link at entry became the focus.
+2 initiativesNew high-return initiatives identified after the People gap closed.
"We were proud of our stack. The diagnosis showed our pride was warranted — and irrelevant. The bottleneck was people, and we'd been underinvesting there for years." — CTO, anonymized
The pattern

What the three cases share.

  • Stack was never the bottleneck. Workflow and People were — every time.
  • The entry score's weakest dimension predicted where the returns were.
  • Systems we built were operable by client teams without us within 90 days — or we consider it failed delivery.
  • No client was steered toward a vendor we profit from. The neutrality is structural, not performative.
Next step

Find your weakest link.找到你最弱的一环——回报就在那里。

Book a diagnostic How ANOI-6 works