Case Study — Product Design

Solex

2024 to Present


The 1st Consulting-Led AI S&OP Solution for Aggregates

I helped Solei build Solex, a product that turns aggregate demand, sales commitments, production constraints and stockpiles into one real-time plan that sites can actually run. I worked with a subject-matter expert to shape a tool that supports an AI decision engine doing the heavy modeling in the background, so site planners and site leads get clear actions to take — and margin gets captured instead of leaking out in the gaps.

~6x

ROI ACROSS A SITE NETWORK

5%+

MARGIN GAINS PER SITE

~300K

SAVINGS / YEAR PER SITE

The Problem

Solex Plan 1: Sales Demand and Plan 2: Inventory Limit comparison screens

Managing a moving target by spreadsheet

Aggregate producers run dozens of product types across every site, each shaped by different crusher specs, screen specs, crew hours, and schedules. When those real-world constraints aren't visible to the people making sales commitments, sites end up over-producing materials nobody's buying and under-producing the ones they are — a constantly shifting target that's nearly impossible to manage with spreadsheets and phone calls.

Solex takes that nightmare away. It gives sales and production planning a shared, real-time view of demand versus what a site can actually make, closing a communication gap that, until now, ran almost entirely on guesswork and good intentions.

The Approach

Turning a modeling problem into a plan someone can act on

I spent time interviewing our subject matter expert to better understand the production pipeline, the role crushers, belts and screens play in cycling the product to break material down into the specific sizes desired by their customers, and how employee scheduling and sales demand fits into that pipeline — then understanding the algorithm we can use on the backend to optimize production and how that can be communicated to sales and production managers/crusher operators. Taking time to understand the problem and audience affected was key.

Research

Aggregate production pipeline flowchart: feeding hopper, crushers, conveyor belts, and screens OEE example calculation: availability, utilization, and overall efficiency for a shift

Iterate

Building an MVP

Once I aligned with stakeholders on a direction for what an initial deliverable could become, I built an ideal experience and then removed from that initial experience, adjusting scope to be able to build exactly what would add the most value initially — with a roadmap in place to continue to improve upon the nature of the product to satisfy the market need.

Solex Performance Summary dashboard with regional and site-level KPIs Solex Big AI Bets and Opportunities table with projected ROI per site
Solex Production Manager design exploration, from onboarding through sign-in and analysis screens

The Result

Proving the Solex at a real site

We received several outreaches from interested prospective customers and have since tried and tested Solex in the field with an improvement that aligned with our ROI predictions. Solex will only scale from here. Next, we'll flesh out scheduling features and more nuanced equipment parameter manipulation to really dial in the output predictions and allow for mobile applications. I'll continue to build out the design system and use Storybook, and Claude Code to help manage it.

View More Work