01 — Overview & Background
Project Background &
Solution Overview
Goodyear’s Finance team manages massive volumes of data across contracts, transactions, and Annual Operating Plans (AOP). Traditionally maintained as complex Excel sheets, extracting insights required professionals to manually search, build pivot tables, and trace broken formulas.
- Time-consuming manual spreadsheet searches.
- Prone to formula errors and lookup omissions.
- High dependency on advanced Excel expertise.
We introduced an AI-powered chatbot capable of answering contextual financial queries instantly. The AOP Chatbot is embedded directly into Goodyear's dashboard, acting as a persistent copilot across three primary screens to transform static grids into interactive intelligence.
- 1. AOP Dashboard (High-volume grid)
- 2. Contracts & Transactions
- 3. Analytics & Variance Insights
- Conversational Financial Queries via NLP
- Excel Data Interpretation (Reads structured grids)
- Contextual AI embedded in Top Navigation
- Efficient Large Dataset Handling
- Smart Summarization of complex rows
- Drill-Down Support with contextual memory
- Automated Variance & Forecast Analysis
02 — Problem, Objectives & Scope
Problem Statement &
Strategic Value
"The Finance team relied heavily on complex Excel spreadsheets to manage AOP data. Extracting insights required manual filtering, pivot tables, and formula checks. Additionally, contract and transaction information was stored separately, making cross-referencing inefficient."
- Finance workflows were strictly bound to Excel.
- Complex datasets increased cognitive overload.
- Limited support for structured decision-making.
- Embed AI globally into the top navigation bar.
- Preserve the familiar Excel-like grid interface.
- Build structured AI prompts for financial data.
03 — User Research & Audience
Personas &
Pain Points Matrix
Pain Points Matrix
Key Research Insights & Design Decisions
| Discovered Issue | Research Insight | Design Decision Implemented |
|---|---|---|
| Complex Excel workflows | Teams rely heavily on formulas and pivot tables. | Integrate conversational AI to automate insight extraction. |
| Fragmented financial data | Contracts and AOP stored in separate systems. | Unify screens with a persistent, shared chatbot layer. |
| Long analysis cycles | Manual filtering slowed down reporting. | Enable real-time AI summaries and instant drill-downs. |
| Low scalability | Excel-based processes don't scale globally. | Build structured backend integration (React + LLM). |
04 — Brand Guidelines & Design System
Enterprise Standards &
Constraints
- Built strictly using Goodyear’s enterprise UI library.
- Designed for a dense, desktop-first experience.
- Integrated within secure, authenticated environments.
- Followed WCAG accessibility and compliance rules.
- Risk: Handling extremely large-scale financial datasets without lag.
- Risk: Maintaining absolute LLM response accuracy (hallucination prevention).
- Opp: Transform static grids into dynamic intelligence tools.
- Opp: Standardize variance explanations across global regions.
05 — Core Interfaces
The 3-Screen
Financial Workspace
The application divides complex workflows into three distinct views, all unified by the persistent Copilot in the top navigation. Click the tabs below to explore the interfaces.
Displays the Annual Operating Plan in a high-density, Excel-like grid structure for familiarity, but connected directly to the live ERP backend.
| Line Item ID | Category | FY24 Forecast | FY24 Actuals | Variance | Status |
|---|---|---|---|---|---|
AOP-MKT-301 | Global Marketing | $12.4M | $14.15M | +$1.75M | Over Budget |
AOP-RD-104 | Tire Tech R&D | $28.5M | $27.80M | -$0.70M | Favorable |
AOP-LOG-502 | Logistics West | $45.0M | $48.20M | +$3.20M | Review Req |
Unifies structured contract records and historical transaction data, previously siloed in a separate procurement system.
| Contract ID | Vendor Partner | Effective Period | Total Value | Compliance |
|---|---|---|---|---|
CTR-24-891 | Synthetic Rubber Supplies | Jan 24 – Dec 26 | $4.5M | Verified |
CTR-24-402 | Global IT Infrastructure | Mar 24 – Mar 25 | $1.2M | Verified |
CTR-24-118 | European Fleet Transport | Jul 24 – Dec 25 | $8.9M | Pending |
Visualizes financial trends, but relies on the AI layer to generate natural-language explanations of the variance drivers beneath the charts.
Region West actual revenue ($30.0M) fell -$5.0M (-14.2%) below forecast ($35.0M) primarily due to delayed fleet tire delivery shipments in August and increased raw material logistics costs.
06 — Live Demo Playground
Interactive AI
Copilot Simulator
Experience the conversational UI. Select a preset query below to see how the AOP Copilot processes financial questions and returns structured insights directly in the chat panel.
07 — Design Process & Architecture
Workflow &
Execution
- ✓ Defined core user query types based on frequent Finance requests.
- ✓ Designed persistent top-nav chatbot placement and interaction behavior.
- ✓ Created the Excel-like grid structure for high-density readability.
- ✓ Established the 3-screen information hierarchy to separate concerns.
- ✓ Fully integrated the Goodyear Design System components.
- ✓ Iterated on feedback with Finance stakeholders to ensure data trust.
- Chatbot Prompt Architecture: Defined guardrails for LLM responses.
- Data Schema Mapping: Connected UI elements to backend AOP structures.
- API Flow Diagrams: Mapped data requests between the UI and ERP.
- Finance Use Case Mapping: Documented the 20 most common variance questions.
08 — Success Metrics & Impact
Business Outcomes &
Final Metrics
"The Goodyear AOP Chatbot successfully transformed static financial spreadsheets into an intelligent, conversational assistant — empowering finance teams to move from manual data navigation to instant AI-driven insight generation."