1 · Standardize intake
Use a short required-field checklist and a single owner for each request. This is usually the lowest-risk first change and may not require AI.
Effort: Low · Confidence: To validate · Risk: Low
This page shows the structure and decision quality of an OERSI diagnostic. It is a fictional workflow with synthetic inputs, not a customer engagement, benchmark, measured result or case study.
Example team: a small software services team. Workflow: receive an RFP or project enquiry, qualify it, prepare a technical brief and draft a proposal for human review.
| Step | Owner / action | Typical failure to check |
|---|---|---|
| 1. Intake | Sales coordinator reads the request and copies details into notes. | Required fields are missing or duplicated. |
| 2. Qualification | Project manager checks fit, deadlines and dependencies. | Questions are sent in separate threads; ownership is unclear. |
| 3. Technical review | Technical reviewer estimates feasibility and effort. | Review waits in an inbox without a visible due date. |
| 4. Proposal draft | Coordinator assembles a brief and a proposal from templates. | Information is re-entered and versions diverge. |
| 5. Follow-up | Owner manually records the next action. | Follow-ups are missed or lack context. |
These failure modes are hypotheses to validate with the actual process owner, not observed facts about any real company.
| Input | Synthetic example | How it is treated |
|---|---|---|
| Requests per week | 12 | Hypothetical input |
| Manual preparation per request | 45 minutes | Hypothetical input |
| Calculated monthly preparation time | 39 hours | 12 × 45 ÷ 60 × 52 ÷ 12 |
| Cash savings | Not established | Requires measured time released, realizability and customer cost data |
The 39-hour figure is only the arithmetic result of the synthetic assumptions. It is not a promised saving or a claim that 39 hours can be eliminated. A real diagnostic uses customer-supplied volumes, sample periods and time measurements; estimates remain labelled as estimates.
Use a short required-field checklist and a single owner for each request. This is usually the lowest-risk first change and may not require AI.
Effort: Low · Confidence: To validate · Risk: Low
Convert approved request details into a consistent project brief and missing-information list. A person verifies scope, estimates and commitments before use.
Effort: Medium · Confidence: To validate · Risk: Medium
Create a task in the team's existing project board when a brief is accepted. Avoid adding a new CRM unless the current tools cannot support the workflow.
Effort: Medium · Confidence: Depends on integration access · Risk: Medium
Controls: least-privilege integration credentials, minimum necessary data, restricted retention, audit trail, retry limits, manual fallback and a documented rollback. No automated proposal sending or customer commitment.
Example recommendation: validate intake consistency first, then pilot structured brief drafting in an isolated test environment. Do not automate pricing, contractual commitments or external messages in the first iteration.
Not included in this preview: customer-specific analysis, production integration, security certification, guaranteed savings or an implementation commitment.