Need a Python developer to create a microservice based on LLMs

Unpack and sort the database in XL format by cities Unpack and sort the database in XL format by cities Unpack and sort the database in XL format by cities

About the Company Trading company. We work with a product group of more than 2000 items across different categories.Current Situation Currently, the nomenclature is maintained in Google Sheets — data is consolidated by tabs (categories). Tab Structure: Product name Price groups: cost price, wholesale, retail Characteristics: weight, quantity per package, etc. Important: the number of columns varies for different product categories, as they have different characteristics.Why the Current Solution is Inadequate Google Sheets does not allow setting access rights at the level of individual columns. We need to: Grant users rights to view certain columns (for example, only cost prices) Grant rights to edit certain columns (for example, retail prices) While restricting access to other columns in the same tabWhat Needs to Be DoneMain Requirements Flexible Rights Management System Access at the level of individual columns (read/write) Assignment of rights by roles or users Management of rights without the involvement of programmers Support for Different Data Structures Different product categories have different sets of characteristics Adding new columns/characteristics without programming Independence from Developers Administration by internal staff Adding categories, columns, users — through the interface Integration with ERP Exporting current prices to our ERP system Export or automatic integration via API Data Analysis Using AI (preferably) Ability to analyze the entire nomenclature list Enrichment, verification, recommendations — if you have ideas, please describeExpected Result A working solution in which: The nomenclature is structured by categories with different sets of characteristics Column rights are flexibly configured (view/edit) Data is exported to ERP The team can manage the system independentlyWhat We Need from You When Responding Describe in general terms how you envision the solution: What tool/platform do you propose

A deep technical verification of three PDF files for authenticity and possible signs of editing or forgery is required. Not only a visual assessment of the documents is needed. The performer must have a good understanding of the internal structure of PDF files and be able to analyze: file metadata; PDF structure and individual objects; creation history and possible editing; software used; embedded fonts, images, layers, and other elements; possible signs of re-saving, conversion, modifications, or backdating of the document; any technical discrepancies that may indicate manipulation of the files. Based on the verification results, a clear written conclusion regarding each file must be provided, indicating the identified signs, risks, and limitations of the verification. We are considering specialists who have practical experience in digital forensics, PDF document analysis, metadata, or verifying electronic files for authenticity. In your response, please briefly describe your experience, methods, and tools that you use for such verification.

General information It is necessary to develop a simple minimalist web system, the main purpose of which is to maintain a client database, create appointments for visits, and automate the process of confirming visits via SMS, sending one-time links through the API from the service itself. The project is being developed in stages. In the first stage, it is necessary to implement only the basic functionality (MVP) so that the system can be used in real work. After launch and testing, it will gradually be expanded with new modules.Main functionality of the first stage User authorization; Client database; Creating and editing appointments; List of appointments (or a simple calendar); Switching between points of sale; Integration with the SMS operator via API; Sending SMS with any text or link for visit confirmation; Confirmation or cancellation of the visit by the client via a one-time link; Displaying the confirmation status directly next to the client's appointment. At the initial stage, instead of a full calendar, the use of a simple list of appointments by days is allowed. Each day should contain a chronological list of bookings indicating the time, client's name, service, employee, and confirmation status. Later, this list can be replaced with a full calendar without changing the system structure. The system must have the ability to switch between points of sale. Each point of sale has its own list of appointments (or calendar), but they all use a common client database.

Copying the database and front for the database (Currently, the front is partially copied and some characteristics are incorrect. The parser filled the database with its own characteristics) Synchronizing the parser with the database and the characteristics of the database Fixing the bug in the parser agent Cropping photos proportionally to the watermark that comes from the parser Working with the buffer with the new database and tools with the objects of this database Fixing minor bugs: for example, text in pop-ups, or small issues on the front, bugs that appear during task execution Connecting a new storage for media storage Logs have stopped writing to the financial operations table, which affects the balance and tariff plan pages (payment of the tariff, account replenishment (successful/not)) Check the payment functionality again and then replace the test billing API from Monobank with the real one Refine access rights for the database administrator user There are currently hanging processes that load the server; this needs to be checked The logic of comments and notes is currently working incorrectly; it needs to be such that for the same object in 2 databases - comments are tied to each database, not to the object, the source of which is the object of this database (I can explain in more detail in private) Fix the sorting in the databases correctly (Currently from the first added to the last. It needs to be the opposite; the sorting functionality is created) Check and correct the accuracy of filters in the databases, possibly add restrictions on filtering with large data volumes