Finance Control Tower | Gregor Smith

The Challenge

Defining the Forecasting Gap

Across finance teams, forecasting was rebuilt manually each cycle in Excel and only submitted into a system at the end. Data lived across disconnected sources, overrides happened through informal handoffs, and teams spent more time reconciling variance than improving forecast quality.

Problem: Financial forecasting teams struggle with fragmented tools, lack of real-time performance visibility, manual data entry errors, limited AI transparency, and difficulty tracking forecast changes.

Role

Senior Product Designer

Timeline

2025-2026

Objective

Centralize forecasting workflow into a unified, transparent, AI-augmented platform

Scale

Weekly forecasting across divisions, product groups, fiscal periods, and adjustment layers

Leadership

Aligned Finance, Product, Data Science, and Engineering on forecasting logic, AI transparency, and audit controls

Tools

Figma Make, FigJam, User Interviews, Usability Testing

User Research

Who We Are Designing For

Primary User Profile

Finance partners responsible for updating forecasts across divisions, product groups, and fiscal periods-balancing trend-based logic with constant business-context overrides.

Goal: Create accurate forecasts quickly and explain variance drivers to leaders.
Pain Point: Manual rebuilds, fragmented inputs, and untracked overrides reduce confidence and slow decision-making.
Need: One workspace with baseline generation, spreadsheet-speed editing, live accuracy metrics, and audit-ready change history.

Finance Control Tower user personas for forecasting stakeholders

Infrastructure

The Ecosystem

The Control Tower orchestrates operational inputs, AI-generated baselines, human adjustment layers, and performance monitoring in one system. Instead of moving between spreadsheets, dashboards, and submission tools, teams work in a single place to create, explain, and govern forecasts end-to-end.

Finance Control Tower forecast editor showing baseline controls and weekly forecast table

The Finance Control Tower consolidates baseline creation, structured overrides, explainable AI reasoning, and performance visibility into a single operational workspace.

Architecture

Structural Organization

Solution: A unified forecasting control tower with Excel-like editing, real-time performance metrics, explainable AI recommendations, and comprehensive audit trails-delivered in a single, accessible interface.

User Centric Views

Finance Control Tower Forecast Insight view

The Journey

Phases of the Design Process

01

Discovery

Mapped current forecasting workflows across teams, surfaced where manual rebuilds and reconciliation happened, and defined trust requirements for AI-assisted baselines.

02

Exploration

Finance teams needed stronger governance without losing the speed and familiarity of Excel, so I retained spreadsheet-style editing while adding 4-5-4 calendar navigation, structured adjustment layers, and hierarchical accuracy views.

03

Validation

Validation showed that a recommendation alone was not enough to establish trust. That shifted the design toward visible drivers, paste validation, bulk-update safeguards, and audit history that made AI-assisted changes explainable and reversible.

Key Feature

Agent Summarizer (Explainable AI)

Forecasting teams wanted AI to help them start faster, but only if they could see what changed, why it changed, and maintain full override control. Rather than present a single black-box forecast, I designed the Agent Summarizer to expose its drivers, conflicts, and differences from manual edits. The result keeps decision authority with the finance partner while still accelerating the starting point.

01 / Variable Attribution

Breaks down which inputs influenced the recommendation-run rate, historical comparisons, and contextual signals-so users can understand drivers, not just outputs.

02 / What Changed + Conflicts

Highlights differences between the AI baseline and manual edits, flags anomalies, and captures rationale so exceptions can be reviewed without backtracking through files.

Finance Control Tower agent summarizer showing explainable forecast drivers and changes

Technical Rigor

Design Standards for Accuracy at Scale

Designed for high data density without sacrificing clarity, validation, or governance. Forecasting tables stay fast to scan and safe to edit, while every change remains traceable across users, time periods, and adjustment layers.

Data Integrity

Excel paste validation and formula-based bulk updates reduce manual entry errors while preserving structured inputs.

Compliance & Audit

Git-style diff history and versioning make it easy to review what changed, who changed it, and why-across cycles and stakeholders.

Finance Control Tower enhanced summary view with high-density forecast details

Collaboration

Consensus Building

Partnered with Product, Data Science, Engineering, and Finance stakeholders to align on forecasting logic, AI transparency requirements, and governance-ready change tracking. The design balanced spreadsheet-speed workflows with the controls needed for audit and shared accountability.

Finance Control Tower baseline selection view comparing forecast options

The Outcome

Business & User Impact

1 Workflow

Consolidated baseline creation, spreadsheet-speed editing, performance monitoring, structured overrides, and audit history into one workspace for weekly planning cycles.

4 Disciplines

Established shared forecasting logic and AI transparency requirements across Finance, Product, Data Science, and Engineering.

3 Trust Controls

Combined explainable recommendations, human override control, and versioned change history so every forecast adjustment remained reviewable.

Back to Work