Practical Work in Information Science "Creating BD (Data Base)"

We are looking for an experienced Shopify developer to create a full-fledged online store from scratch with mandatory integration with Amazon. What needs to be done: Integration with Amazon as the foundation of the project: synchronization of products, inventory, and orders between Shopify and the seller account on Amazon. The store setup is built with this integration in mind from the very beginning. Store setup: theme customization, catalog structure, product cards, cart, checkout — everything is aligned with the logic of integration with Amazon. Technical optimization for AI visibility (GEO): structured data, llms.txt file, content optimization for AI assistant responses (ChatGPT, Perplexity, AI Overviews, etc.), proper semantic markup of pages. Basic SEO setup: meta tags, sitemap, loading speed, mobile responsiveness. Connecting payment systems. Installing necessary applications (reviews, email marketing, analytics, etc.). Testing the store before launch.

Need help rethinking and building a clean relational schema for an internal operational system on Airtable. The current database is already in use by the team, but it has grown organically: the structure is partially flat, some tables/views are actively used, while others are hardly used, and documents currently rather "pull" information into Airtable, although ideally Airtable should become the source of truth and generate the necessary operational artifacts. What is needed is not just help with forms or automations, but structural work: analyzing the current logic, designing a relational schema, creating a clean new database/instance, documentation, and recommendations for migrating to the new schema. And potentially even implementation. System context: There are two main directions: Feedback system A system for collecting and processing signals from external advisors, focus groups, internal teams, and stakeholders. It should support: signal log; advisor tracking; feedback intake; synthesis dashboard; status/gates/recommendations; close-the-loop logic. Product catalog The product catalog currently has about 400 entries, approximately 100 of which are active/live. It is necessary to better structure products, statuses, relationships, data sources, lifecycle/gate logic, and reporting. Current state: Airtable already exists and is used by the team. The current data accuracy is about 80%. The schema is mostly flat, not relational. Only the first few tables/tabs are actively used: Some validation gates, tasks/decisions, and dashboard views are currently hardly used. There is no PII. Any work with the live base must be done with maximum caution: no changes in production without agreement. The ideal approach is to analyze the structure and build a new clean base separately in Sandbox. What needs to be done: Phase 1 — Audit & architecture recommendation Look at the current structure of Airtable. Determine which tables/fields/views are actually needed, which are duplicated, which can be merged or removed. Propose a relational data model. Describe the main entities, relationships, primary keys, linked records, lookup/rollup logic. Propose a clean schema for the Project + Product catalog. Determine which dashboards/interfaces are needed for different users. Prepare a short architecture memo or schema map. Phase 2 — Build clean Airtable base Create a new clean Airtable base/instance. Set up tables, fields, linked records, views, basic interfaces. Set up basic dashboard/reporting logic. Prepare a migration map: how to transfer data from the old structure to the new one. Document the structure so that the team can maintain it after handover. If necessary, propose an automation strategy, but without excessive complexity. Expected deliverables: Airtable schema map. A new clean Airtable base. Tables + relationships + key fields. Views/interfaces for main users. Migration recommendation. Short documentation for the team. Governance recommendations: who enters data, who approves changes, which fields are mandatory, how to avoid duplication. Optionally: 1 short handoff call / Loom walkthrough. Important security/access rules: No PII. No dangerous tokens or personal integrations. AI/MCP/ChatGPT/Claude cannot be connected to live Airtable. If access to the existing base is needed, it must be read-only or through export/screenshots/structural description. Any records/changes are made only in the new test/clean base, not in live production. Who I am looking for: The ideal candidate has experience in: Airtable base architecture; relational schema design; Airtable interfaces, views, forms, automations; migration/cleanup of messy Airtable bases; product operations / CRM / workflow systems; documentation and handoff. In your response, please indicate: Examples of Airtable bases or systems you have built. Whether you have experience transforming a flat Airtable structure into a relational schema. How you would approach audit → schema design → build. An estimated number of hours for Phase 1. Your hourly rate or fixed-price proposal for the first phase. Whether you are willing to work under an NDA. Work format: It is preferable to start as soon as possible. Initially, a small paid discovery/audit can be done, after which we can move to the full build.

Need to update the offer agreement in the footer to a new one, the company has changed its regulations Task will take 20 minutes

Technical Assignment: Tool for Reconciling Accounts with Bank, Cards, and AccountantGeneral Goal A local tool (script/small application in Python) is needed, which is manually run once every 1-2 months on my computer and performs reconciliation between: Invoices I issued to clients and received from suppliers (PDF files and photos of receipts stored in Gmail) Bank statement (Excel) Credit card expenses (Excel, may have several files — multiple cards) Excel file from the accountant, in which he enters invoices for reporting to tax and VATWhat the Tool Should Do1. Collect Invoices from Email (Gmail) Connection via Gmail API (OAuth) Search for emails over a specified period (e.g., the last 2 months) with attachments Download attachments: PDF files and images (photos of receipts/invoices) Recognize data from each invoice: date, amount, name of supplier/client, invoice number (if available) For PDF — extract text directly For images — text recognition (OCR), including text in Hebrew2. Read Excel Files Bank statement (income and expenses) One or more files of credit card expenses Excel from the accountant with a list of entered invoices (it contains only the amount + date + name of the counterparty, there is no unique invoice number for matching)3. Matching Logic Match records between sources based on the combination: amount + date (with an acceptable deviation) + name of the counterparty (fuzzy text comparison, as names may differ slightly).4. Checks to Perform A. Are invoices issued for all income For each incoming money in the bank — check if there is a corresponding invoice that I issued. If not — mark as "income without issued invoice". B. Are invoices saved for all recognized expenses For each expense in the bank/on cards that relates to the category of recognized expenses — check if there is a saved invoice. If not — mark as "expense without saved invoice". C. Reconciliation with the accountant's list Compare the list of invoices I collected (from email) with the list that the accountant entered and sent to tax/VAT. Mark: Invoices that I collected but are not in the accountant's file (possibly not transferred or not entered) Records in the accountant's file for which no corresponding invoice was found in my collection D. Payroll Check Check that the amount stated in the payroll matches the amount that was actually debited/credited in the bank.5. Final Report The result — one Excel file with several sheets: Income without issued invoice Expenses without saved invoice Invoices missing from the accountant Accountant's records without corresponding invoice Payroll reconciliation with the bank Discrepancies/problematic lines should be visually highlighted (in color).Technical Details Language: Python Excel: pandas + openpyxl PDF: text extraction (pdfplumber/PyMuPDF) OCR for images: with support for Hebrew (preferably use Claude API with image support instead of classic Tesseract, for better accuracy with Hebrew) Gmail API: google-api-python-client, OAuth (a project will need to be created in Google Cloud Console — this will be done by the client/performer together) Result is saved as an Excel file locally on the computer Interface: command line (CLI) is sufficient, no need for a web interfaceSubmission Format Source code of the project Brief instruction on how to run (including how to connect your Gmail and where to place the input Excel files) Instructions for creating credentials in Google Cloud Console, if not done in advanceAdditional I am ready to provide examples of the structure of Excel files (bank, cards, accountant's file) and examples of invoices (PDF and photos) for testing.