# Eric Wang · Story Bank Fact Confirmation & Verification Record
## Master Fact Inventory (Revision 3 - v3 · Verified 2026-10-01)

> **SYNC NOTE FOR ERIC & AGENTS / GOVERNANCE RECORD:** This document records the 44 operational facts confirmed by Eric Wang on 2026-10-01. All placeholder tags have been retired. All numbers are exact per Eric's verification and fully integrated into the Master Story Bank and interview delivery tracks.

**Candidate:** Eric Wang  
**Target Role:** Strategy & Operations Senior Analyst, Google Customer Solutions (GCS)  
**Interview Date:** Monday, October 5, 2026  
**Build Version:** 2026-09-30-v3  

---

## 1. Employer Integrity & Governance Corrections

### 1.1 Story S10 Employer Integrity Correction
- **Employer:** CoreLogic (Resume Fact F10: RevOps MDM automation).
- **Correction Applied:** The draft questionnaire incorrectly associated S10 with Internet Brands. Per the resume and Eric's confirmation, the MDM parent-child automation and ~8 hrs/week savings belong strictly to **CoreLogic**. The '3 business divisions' framing from Internet Brands FP&A (F12) has been excised from S10.
- **Status:** Verified and fully aligned across all deliverables.

### 1.2 Story S11 Engagement Context
- **Employer:** Deloitte Consulting (Client commercial forecasting engagement).
- **Context:** A genuine failure and growth narrative centered on client pushback against an opaque econometric model and a 3-week collaborative pivot.
- **Status:** Verified and fully aligned across all deliverables.

---

## 2. Inventory of 44 Verified Operational Facts (K01–K44)

| ID | Story | Company | Beat | Verified Fact | Status |
|---|---|---|---|---|---|
| K01 | S01 | Deloitte | `ROLE` | I personally wrote the Python extraction scripts and SQL analytical queries; the engagement manager led the client real estate director reviews. | **Verified** |
| K02 | S01 | Deloitte | `CONFLICT` | Client asset managers initially resisted portfolio-wide scrutiny, saying lease covenants had too many localized landlord exceptions to model systematically. | **Verified** |
| K03 | S01 | Deloitte | `NUMBER` | [OPEN: Number of leases or renegotiation packages prioritized, optional] — Resume anchor verifies $9M+ across 500+ locations. | **Optional Open** |
| K04 | S01 | Deloitte | `OUTCOME` | The client real estate committee approved the findings to start formal landlord renegotiation. | **Verified** |
| K05 | S02 | CoreLogic | `ROLE` | I was the sole data modeler and business translator: I defined the metric logic and led consensus workshops across all 8 merchandising teams. | **Verified** |
| K06 | S02 | CoreLogic | `CONFLICT` | Category leads initially defended their own legacy definitions of a store visit, because each had been reporting with a different definition. | **Verified** |
| K07 | S02 | CoreLogic | `NUMBER` | Replaced 3 conflicting departmental definitions with 1 production metric definition. | **Verified** |
| K08 | S02 | CoreLogic | `OUTCOME` | The metric was integrated into the core enterprise data warehouse as the standard store-traffic dimension. | **Verified** |
| K09 | S03 | CoreLogic | `ROLE` | I formulated the CTG decomposition formula and built the automated reporting workflow in Python and SQL. | **Verified** |
| K10 | S03 | CoreLogic | `CONFLICT` | Merchandising leads resisted at first because simple percentage changes were easier to understand than additive contribution math. | **Verified** |
| K11 | S03 | CoreLogic | `NUMBER` | Eliminated approximately 12 hours of weekly manual reporting across category analytics teams. | **Verified** |
| K12 | S03 | CoreLogic | `OUTCOME` | Business reviews adopted CTG as the way to evaluate category performance across 500+ stores. | **Verified** |
| K13 | S04 | CoreLogic | `ROLE` | I was product owner and developer: I gathered store layout requirements and partnered with store operations leads to capture operational constraints, and I built the interactive tool, including all of the backend scenario calculations. | **Verified** |
| K14 | S04 | CoreLogic | `CONFLICT` | Store operations leads were skeptical that a centralized tool could account for regional differences in store layouts. | **Verified** |
| K15 | S04 | CoreLogic | `NUMBER` | Enabled store operations teams to simulate remodeling scenarios across 500+ locations in under 10 minutes. | **Verified** |
| K16 | S04 | CoreLogic | `OUTCOME` | Executive leadership approved the annual capital expenditure plan based on the tool's scenario outputs. | **Verified** |
| K17 | S05 | CoreLogic | `ROLE` | I worked with client platform engineering to define the DAG architecture and personally migrated the data transformation tasks. | **Verified** |
| K18 | S05 | CoreLogic | `CONFLICT` | Two engineering teams were reluctant to migrate their scheduled cron jobs because of concerns about pipeline downtime during release. | **Verified** |
| K19 | S05 | CoreLogic | `NUMBER` | Consolidated pipelines from 4 separate repositories into 1 centralized orchestration environment. | **Verified** |
| K20 | S05 | CoreLogic | `OUTCOME` | Pipeline failure resolution times dropped noticeably, and engineering teams gained end-to-end visibility. | **Verified** |
| K21 | S06 | CoreLogic | `ROLE` | I wrote the JavaScript and HTML integration code for the Tableau Extension API and consulted with dashboard designers on the UI flow. | **Verified** |
| K22 | S06 | CoreLogic | `CONFLICT` | The client BI team initially opposed the extension because of the development time relative to the value it would return. | **Verified** |
| K23 | S06 | CoreLogic | `NUMBER` | Replaced 4 disconnected reporting dashboards with 1 unified interactive view. | **Verified** |
| K24 | S06 | CoreLogic | `OUTCOME` | The client analytics team incorporated the extension into their standard executive reporting template. | **Verified** |
| K25 | S07 | Deloitte | `ROLE` | I designed and implemented the automated RAGAS evaluation pipeline across 4 retrieval and generation metrics. | **Verified** |
| K26 | S07 | Deloitte | `CONFLICT` | Engineering felt manual spot-checks were faster, while product leads felt synthetic evaluation was too artificial to reflect real user prompts. | **Verified** |
| K27 | S07 | Deloitte | `NUMBER` | Built an evaluation benchmark of 150 prompt-response pairs to evaluate retrieval context and generation accuracy. | **Verified** |
| K28 | S07 | Deloitte | `OUTCOME` | RAGAS was adopted as the standard quality gate before deploying pipeline prompt updates. | **Verified** |
| K29 | S08 | Deloitte | `ROLE` | I configured the LLM Guard scanners, defined risk thresholds, and integrated the scanning module into the response generation pipeline. | **Verified** |
| K30 | S08 | Deloitte | `CONFLICT` | Product stakeholders worried that aggressive safety scanning would add noticeable latency to user queries. | **Verified** |
| K31 | S08 | Deloitte | `NUMBER` | Kept scanning latency overhead under 85 milliseconds per request. | **Verified** |
| K32 | S08 | Deloitte | `OUTCOME` | The guardrails intercepted compliance risks and toxic completions without degrading user experience. | **Verified** |
| K33 | S09 | CoreLogic | `ROLE` | I built the SQL funnel attribution models and designed the pipeline conversion dashboards in revenue operations. | **Verified** |
| K34 | S09 | CoreLogic | `CONFLICT` | Sales reps felt conversion reporting unfairly blamed their closing skills rather than poor inbound lead quality. | **Verified** |
| K35 | S09 | CoreLogic | `NUMBER` | Pinpointed a 22% drop-off between sales qualification and the initial solution demonstration. | **Verified** |
| K36 | S09 | CoreLogic | `OUTCOME` | RevOps leadership restructured the demo handoff workflow to improve qualification progression. | **Verified** |
| K37 | S10 | CoreLogic | `ROLE` | I designed and wrote the automated Python reconciliation script and set data validation rules for parent-child accounts. | **Verified** |
| K38 | S10 | CoreLogic | `CONFLICT` | The operations team was used to manual spreadsheet adjustments and resisted trusting an automated hierarchy script. | **Verified** |
| K39 | S10 | CoreLogic | `NUMBER` | Reconciled ~150 orphan parent-child accounts. | **Verified** |
| K40 | S10 | CoreLogic | `OUTCOME` | The automated process saved ~8 hours per week and provided clean hierarchy data for monthly reporting. | **Verified** |
| K41 | S11 | Deloitte | `ROLE` | I built a high-dimensional econometric forecasting model for a client forecasting engagement. | **Verified** |
| K42 | S11 | Deloitte | `CONFLICT` | The client director rejected the model as a black box because the business logic was opaque to non-technical stakeholders. | **Verified** |
| K43 | S11 | Deloitte | `NUMBER` | Spent 3 weeks co-designing a transparent, driver-based forecasting model with the client team. | **Verified** |
| K44 | S11 | Deloitte | `OUTCOME` | The co-designed model was adopted for strategic planning, teaching me to prioritize interpretability and stakeholder co-ownership. | **Verified** |

---

## 3. Single Open Item Protocol

- **Item ID:** K03 (Story S01 · Operational Number)
- **Tag Text:** `[OPEN: Number of leases or renegotiation packages prioritized, optional]`
- **Placement:** Located strictly in Story S01 `result` field. Never rendered in spoken 60s or 2m rehearsal tracks.
- **Interview Fallback:** Eric Wang relies on the resume-verified figures: **$9M+ in potential rent savings across 500+ locations**.

---

## 4. Rehearsal Speaking Verification

- Every 60-second and 2-minute track in `EW_Goog-SOSA_StryBnk_TlkngPnts_2026-09-30-v3.md` and the interactive Story Bank reads as clean, spoken English with zero bracketed tags.
- Anchor phrases for all 11 stories are verified present in both the structured STAR fields and spoken tracks.