Creation of semantic core, subcategories of requests

A separate CRM system needs to be developed for the centralized processing of all retail orders from various sales channels (Prom, Rozetka, own websites, and manual orders). Main requirements: A unified order registry from all sources, preserving information about the source and its status. A separate internal CRM status that operates independently of the status on the marketplace or website, with two-way synchronization of statuses where supported. Flexible integration with 1C: for each source, the administrator can separately enable or disable automatic order export. Integration with carriers for automatic creation of shipping documents, saving their numbers in the CRM, and transmitting the shipping document numbers to Prom and Rozetka. Integration with Checkbox for creating receipt projects, using customizable payment method matching, with subsequent fiscalization control. Support for manual order creation (phone, Viber, Telegram, Direct, etc.) with a complete subsequent processing cycle through the CRM. All API keys and integration settings must be managed by the administrator through the CRM interface without the involvement of a programmer. A complete action log and integration logging for system operation control. Main goal The manager must be able to fully process orders in one CRM interface — from receiving the order to creating a document in 1C, shipping document, receipt project, and completing the order, without the need to access the cabinets of Prom, Rozetka, websites, carriers, or Checkbox.

We have synchronization of two data sources with the CRM. The system as a whole is stable. However, there remains a part that becomes desynchronized. The main hypothesis is that in one of the sources, the data changes after the initial synchronization, and these changes are not always pulled into the CRM. + the data in the two sources have different names/identifiers even though they represent the same entities.

Task: deploy an LLM service that knows all the company's documentation and answers questions from the sales department managers. Current situation: the client has independently assembled a prototype (a separate project with uploaded company information, hosted on a server), but the information from the database is not transmitted to the model — likely, there is an issue with the API. We will provide the code and access. The first step is an audit: fix the existing setup or justifiably rebuild from scratch. Required functionality: Upload all company documentation: description of each service, regulations, FAQ, pricing (all materials will be provided). Answers strictly based on the uploaded documents (RAG). The model does not invent facts; if the answer is not in the database — it honestly informs about it. Access for managers via a link (web interface), with authorization. Scenarios: the manager asks any question about the company's work; inserts the client's question "as is" and receives a ready answer for sending; finds the necessary regulation/report by request. Knowledge base updates without a developer (uploading files through the interface or a connected folder). English language. History of requests for quality control. Technical expectations: LLM via API (Claude/OpenAI — propose with a cost calculation for tokens), RAG pipeline (vector database, embeddings), hosting on our server or in the cloud, HTTPS. The architecture should allow for future connection of the assistant to the analytical data warehouse (parallel project). In the response, indicate: examples of similar RAG projects, stack, timeline, cost of work, and estimated monthly ownership cost (tokens + hosting).

Hello There is a small department where patients are admitted. It is necessary to develop a simple "database" where consumables, medications, and others can be entered, along with incoming and outgoing records, as well as assigning usage to a specific patient. Required functionality Directories Medications Consumables (manual entry option) Patients (ability to continuously add new ones and delete those who have been discharged) Employees (ability to modify)Movement Accounting A single table (movement log) where everything is registered Incoming and write-off Manual correction option When writing off, the nurse must select: date; patient; medication or material; quantity; her full name. The system automatically reduces the balance. When receiving, the head nurse must be able to enter: Delivery date; Medication/material; Remaining quantity; Source; Expiration dateAutomatic Functionality automatic balance calculation; minimum stock control; expiration date control; low stock warnings; protection against writing off more than is available; drop-down lists; data entry validation; conditional formatting.Reports It is necessary to implement automatic reports: current balances; usage over a period; usage by patient; usage by employee; history of movement of a specific medication.Preferred implementation on Google Sheets; We will fill in the databases ourselves. A template for work is needed. I don't know how much this costs, so let's start with $50, and you can suggest your options, preferably with experience.

It is necessary to calculate statistics in a medical project (including correlations). Additionally, there is a request to create a mathematical model for predicting the development of pathologies based on available data.