A custom operations platform for a federal supply business, designed from screenshots of a manual workflow and developed into a database-driven production system in under 90 days.
Demo recreation of a live production system. Fictional company and sample data shown for confidentiality.
Ace Tomato Supply Co. is a fictional brand created for this portfolio demo. Its photography is AI-generated illustration.
The build began with screenshots of spreadsheets and notes already used by the business, then turned those habits into a structured web application.
The real application manages opportunity research, bid tracking, supplier sourcing, batch RFQs, contract awards, purchase orders, shipments, exports, events, and leadership reporting in one database-driven workflow.
It was designed, specified, directed, reviewed, tested, and deployed through AI-assisted development, with Claude used as the coding partner. The public presentation intentionally reveals the scale and engineering discipline without exposing the real operation.
The mock demonstrates the functional scope of the private production platform without using real customers, suppliers, identifiers, pricing, or live records.
Move opportunities from research to quoting, submission, award, or closeout with due dates and status visibility.
Organize part and stock-number research tied to imported federal parts-catalog data for sourcing analysis.
Maintain manufacturer directories, capability tags, sourcing notes, and historical quote activity.
Select multiple parts and multiple suppliers, then organize one structured request-for-quote workflow.
Track awards, purchase-order milestones, shipments, and delivery status from handoff through completion.
Surface weekly operating summaries, pipeline views, events, and spreadsheet exports for decision support.
Every interface below is a fictional recreation. Names, dates, values, company records, part numbers, and opportunity codes are invented for presentation.
The system was not generated in one pass. It was continuously designed, documented, tested, migrated, backed up, reviewed, and improved.
Start from screenshots of the existing spreadsheet-and-notes workflow.
Translate real business activity into structured pages, data relationships, and rules.
Develop the application with Claude as the AI coding partner under human direction.
Use migrations, automated tests, staging, error-log triage, backups, and rollback copies.
Keep shipping. More than 20,000 lines were added after the first major build phase.
The public portfolio stays at a general technology level and intentionally avoids infrastructure, hosting, database naming, URLs, credentials, or implementation details that identify the live system.
This page is the public-safe presentation. The live application, its data, customers, suppliers, pricing, and operating methods remain confidential.