Chemical and fuel trading and distribution
The system a chemical trading business runs on
A legacy CRM could track deals, but it could not track the thing the business actually lives on: realized margin on physical loads. So we built the platform that does, from the data model up.
- 237K+
- Lines in production
- 100
- Tables modeled
- 1
- System of record
- 24/7
- In production
The challenge
What the old system could not do
Chemical trading is a spread business. You buy product from suppliers, sell it to customers, and the profit is the margin minus freight. Most CRMs cannot model that. They track deals, not physical loads of product with real, realized economics.
The business was running on an off-the-shelf CRM that could not represent loads, margins, or the industry workflows the team depended on. Every honest answer about how the business was doing meant building another spreadsheet by hand.
What we built
Chemical and Fuel Trading ERP
- 01
The deal lifecycle, modeled correctly
Prospects flow into opportunities, then into the deal blueprint of which supplier, which customer, which product, at what price, and finally into loads, the actual executions. Loads are the single source of truth for realized margin, revenue, freight, and delivered volume, sliced by rep, product, customer, and business entity.
- 02
Industry workflows no off-the-shelf CRM has
A product and specification catalog, a samples and shipment pipeline, typed laboratory result tables for quality control, and AI-assisted preparation of regulatory safety documents. The parts of the business generic software refuses to handle.
- 03
AI built into the core, not bolted on
A governed interface exposes the system of record to AI agents with scoped, revocable read and write access. Every agent action is attributed to a named rep, written to an audit ledger, and reversible. A rep can talk to an agent by phone and have it read and update records hands-free.
- 04
The full operating surface on top
Accounts, contacts, correspondence, calls, meetings and recurring tasks, documents, a configurable dashboard, a custom report builder with drill-downs, and an automation engine that fires on field changes, record creation, and date proximity.
The outcome
What changed once it shipped
- Every dollar of realized margin is attributed to the right rep, product, customer, and entity automatically, at load level.
- "How are we actually doing?" went from a spreadsheet exercise to a live report.
- Reps work alongside AI agents that read and write the system of record, with every action attributed and reversible.
- Features the old vendor would never have justified building now ship on demand.
Built with
The stack underneath
- TypeScript end to end
- React front end
- Managed relational database
- Serverless hosting
- Governed agent interface
- Automated test suite
Ready for the system your business runs on?
We start by mapping how the operation actually works, then build the platform that fits it. If your industry is still open, the first conversation costs an hour.