DealFW Scanning Platform↗
From an unknown item to a reviewed auction lot.
1. Auction selection — choose an open auction with available lot capacity.
2. Photo capture — upload a structured set of product images using bounded-concurrency storage uploads.
3. Product identification — combine Gemini-assisted image analysis with a searchable product catalog, while keeping the employee responsible for the final match.
4. Condition testing — generate an item-specific checklist, capture the employee's test notes or recording, and convert the results into structured condition evidence.
5. Listing review — review and edit the proposed title, description, category, condition, and marketplace data before anything is finalized.
6. Finalization and label — allocate the next SKU and lot number transactionally, write the listing, render a printable label, and return it to the employee.Open it.
Inspect the work.
The internal employee scanning app and admin dashboard are documented separately from the public storefront. This original write-up explains the scan-to-label workflow and its review boundaries.
1. Auction selection — choose an open auction with available lot capacity.
2. Photo capture — upload a structured set of product images using bounded-concurrency storage uploads.
3. Product identification — combine Gemini-assisted image analysis with a searchable product catalog, while keeping the employee responsible for the final match.
4. Condition testing — generate an item-specific checklist, capture the employee's test notes or recording, and convert the results into structured condition evidence.
5. Listing review — review and edit the proposed title, description, category, condition, and marketplace data before anything is finalized.
6. Finalization and label — allocate the next SKU and lot number transactionally, write the listing, render a printable label, and return it to the employee.Excerpt from the published engineering case study. This is workflow documentation, not application source code or a screenshot.
Documentation reviewed September 14, 2026. The scanner itself was not run during this portfolio update; no scanner-specific test result or interface capture is claimed.
A real problem.
A considered solution.
Incoming liquidation stock has inconsistent packaging, incomplete product information, and uncertain condition. Employees need one reliable path from an unidentified item to a reviewed, labeled auction lot.
What I worked on.
I built major parts of the employee scanning and admin platform: photo intake, AI-assisted identification, condition testing, listing review, lot allocation, labels, and operational reporting. My work spans the software and the warehouse operation it supports.
Photo intake
An employee selects an auction and adds product photos. Uploads are bounded so a large batch does not overwhelm the workflow.
A guided explanation of the project architecture. This explorer does not connect to production systems.
Give operators a complete workspace.
The admin dashboard brings together inventory, auctions, pallets, employee access, label reprinting, exports, productivity, and time-entry review. Employees work within a narrower New Scan, My Products, and My Stats interface.
People approve. AI assists.
Gemini helps identify products and structure condition notes. Employees confirm the match, condition, and listing before finalization.
Allocate lots transactionally.
Server-side Firestore transactions allocate SKU and lot numbers, protecting the sequence when employees scan simultaneously.
Keep public and private data separate.
The storefront reads a limited public projection. Staff records, internal inventory fields, and administrative workflows remain behind role-aware access.
An employee scanning workflow and administrative platform connecting product intake, reviewed listings, auction lots, labels, and warehouse operations.
Internal production application. The public engineering write-up documents the design; production source and an interface walkthrough are not included here.
Invoice OS
Arlington, TX · Open to Seattle and US relocation