Build MVP: AI Product Catalog → Zoro NPI Excel Template Automation

Build MVP: AI Product Catalog → Zoro NPI Excel Template Automation Budget: $400 fixed price I am looking for an experienced developer to build a small standalone web application that converts a large product catalog into a completed Zoro.com NPI Excel template. I currently have more than 10,000 SKUs/products in an ERP / ecommerce catalog. Today, transferring this information manually into Zoro’s NPI workbook is very time consuming. The goal is to automate most of this process. What I Will Provide I will provide: * The official Zoro NPI .xlsm template * Example completed Zoro products * Sample ERP/product export * Sample Shopify product export if needed * Product data including SKU, title, description, price, dimensions, weight, images, etc. * OpenAI API key for AI functionality if required The supplied Zoro workbook contains multiple sheets including: * NPI * Images & Assets * Attributes * Variants * Product Selling Features * Item Relationships * LOV/reference values The developer must inspect the supplied workbook and preserve the Zoro-required structure. ⸻ Main MVP Workflow The application should have a simple web interface. Step 1 — Upload Product Data User can upload: * CSV * XLSX The source could be an ERP export, Shopify export, or another product database. The system should not be hardcoded to only one exact input file. ⸻ Step 2 — Map Fields The application should allow columns from the source catalog to be mapped to Zoro fields. Example: SKU → Supplier Stock No. SKU → MFR Model No. Vendor → Manufacturer Name Title → Product Title Description → Product Long Description Cost → Product Cost Price → Retail Price Weight → Ship Pack Weight etc. Mapping should be saved so it does not need to be recreated every time. ⸻ Step 3 — AI Listing Optimization Use the OpenAI API to improve listing information where appropriate. AI should be able to generate or improve: Product Title Maximum according to the Zoro template. Titles should be optimized for: * product type * material * size * thickness * important attributes * relevant buyer search terms AI must NOT invent product specifications. ⸻ Long Description Generate a clean professional product description using the available product information. It should include useful product information such as: * what the product is * material * dimensions * important specifications * common applications * benefits * relevant search terminology Must respect Zoro character limits. ⸻ Product Selling Features Generate up to 10 useful selling features for the Zoro: Product Selling Features sheet. Features should be product-specific and not meaningless generic marketing text. ⸻ Keywords / Search Optimization AI should improve wording so products can be found for relevant commercial/industrial searches. This is not just normal Google SEO. The goal is to create accurate marketplace product content using the terminology a buyer may use when searching for the product. ⸻ Step 4 — Zoro Field Validation Before export, the system should identify: * missing required fields * invalid values * character-limit problems * incomplete dimensions * incomplete shipping information * products requiring review * fields that cannot safely be determined The Zoro workbook contains LOV/reference values. Where appropriate, the application should validate values against these allowed values instead of generating random text. Examples include fields such as: * Unit of Measure * shipping method * country of origin * warranty * regulatory answers * state restrictions * image/asset types AI should never guess regulatory or compliance answers. If data such as Prop 65, hazardous material status, country of origin, EPA information, SDS status, etc. is unavailable, the product should be flagged for human review. ⸻ Step 5 — Review Interface I need a simple dashboard/table showing products. Example: SKU Product Status Missing Fields AI Complete Ready for Zoro User should be able to filter products by: * Ready * Warning * Missing information * AI processing * Error I do NOT need a beautiful enterprise UI. A clean, functional interface is enough. ⸻ Step 6 — Generate Zoro Workbook This is the most important requirement. The application must populate the supplied Zoro NPI template while preserving the workbook structure. It should generate the appropriate data for relevant sheets including: * NPI * Images & Assets * Attributes * Variants where applicable * Product Selling Features * Item Relationships where applicable The original Excel structure, worksheets and required formatting must remain compatible with Zoro. The system should allow me to download the completed Zoro file. It should work with thousands of products, not just 10-20 rows. ⸻ Important Architecture Requirement Although this is currently a standalone application, it may later become a module inside a larger application. Therefore, please do NOT build all business logic directly inside UI components. I want a basic separation such as: Import Layer Reads CSV/XLSX product data. ↓ Normalization Layer Converts different product source formats into one internal product format. ↓ AI Enrichment Layer Generates titles, descriptions and selling features. ↓ Zoro Mapping / Validation Layer Maps normalized product data to Zoro requirements. ↓ Zoro Export Layer Generates the completed Zoro workbook. The UI should call these services rather than contain all transformation logic itself. ⸻ API / Integration Requirement Please create a small internal API so another application can use this system later. For example: POST /api/import POST /api/products/normalize POST /api/products/enrich POST /api/products/validate POST /api/export/zoro Exact endpoint names can differ. The important requirement is that the conversion/export engine can later be called programmatically without depending on the current UI. ⸻ Internal Product Format The application should create a normalized product object/schema internally. For example: * SKU * manufacturer * manufacturer model * brand * title * description * product type * category/taxonomy * cost * retail price * length * width * height * weight * unit of measure * country of origin * lead time * image URLs * attributes * selling features * compliance information * additional metadata This will make it possible to add different input sources later. ⸻ Technology I am open to recommendations. Preferred technologies include: * Next.js / React * Node.js / TypeScript * Python for Excel processing if appropriate * OpenAI API * SQLite/PostgreSQL or another lightweight database You do not need to use all of these. The application should be easy for another developer to understand and continue developing. ⸻ NOT Required for This MVP Please do NOT include the following in the $400 proposal unless very easy to add: * Direct Zoro API submission * Zoro order management * Full ERP API integration * Full Shopify OAuth integration * Inventory synchronization * Customer management * Accounting * Complex user permissions * Billing/subscriptions * Enterprise UI This project is specifically about: Product catalog → AI enhancement → Zoro NPI validation → Zoro Excel export. ⸻ Deliverables For the $400 fixed-price MVP I expect: 1. Working application 2. Source code 3. CSV/XLSX product import 4. Reusable field mapping 5. AI title/description/selling-feature generation 6. Zoro validation 7. Product processing/status dashboard 8. Zoro .xlsm workbook generation 9. Basic internal API/service architecture 10. README with installation instructions 11. .env.example 12. No API keys or passwords hardcoded in source code ⸻ Test Requirement Before the project is considered complete, I will provide a sample catalog. The application must successfully process the catalog and create a Zoro workbook from it. We will first test a small number of products. After that, we must test a larger batch to confirm the architecture can handle thousands of SKUs. ⸻ Very Important Please do not apply if your solution is simply: “I will use ChatGPT to generate an Excel file.” I need an actual reusable application and transformation engine. The system must understand the Zoro workbook structure, map source data correctly, validate required fields, preserve the template, and be designed so the engine can later be integrated with another application. ⸻ When Applying Please answer these 5 questions: 1. Have you previously generated or modified complex .xlsx or .xlsm Excel templates programmatically? Give an example. 2. What technology would you use to preserve and populate an existing .xlsm workbook? 3. How would you separate the import, AI enrichment, validation and Excel export logic? Short explanation only. 4. How would you prevent AI from inventing product specifications or regulatory information? 5. Can you complete the described MVP for the fixed $400 budget? Please start your proposal with: ZORO-MVP Applications without ZORO-MVP at the beginning will not be considered.