Based on the technical brief provided for the implementation of Kommo CRM in 2026, here are the project guidelines and scope, with the removal of information that directly identifies the client: Project Context and Objective The project consists of the complete implementation of Kommo CRM for a physical retail store and wholesaler located in Ciudad del Este, Paraguay. The primary focus is the automation of customer service via WhatsApp for the sale of electronics, Arabic perfumes, and thermal products. 1. Account and User Configuration Connection: Integration of WhatsApp Business with Kommo. Profile: Configuration of name, logo, and language (Portuguese). Team: Configuration of 6 users with specific permissions: 1 Administrator: Full access. 1 Manager: View of all leads, pipelines, and reports. 5 Users: Operational access to leads and conversations. Standard: Every new message must automatically create a lead. 2. Pipeline Structure Two distinct pipelines must be created: WHOLESALE Pipeline — B2B New Lead: Initial service via Bot. Qualifying: Bot collects volume and location data. Qualified Lead: Notification sent to the salesperson. Catalog Sent: Wholesale price list sent. In Negotiation: Active conversation. Proposal Accepted: Awaiting payment. Won: Deal closed. RETAIL Pipeline — B2C New Contact: Basic responses via Bot. Interest Identified: Focus on a specific product. Invited to Store: Address and invitation sent. At the Store: In-person confirmation by the salesperson. Sale Completed: Automated post-sales follow-up. Review Requested: Request for feedback on Maps. 3. Required Custom Fields 9 fields must be created under Settings > Fields > Leads, with names identical to those listed below to prevent Bot errors: Client Type (List). City / State (Text). Product of Interest (Multiple Choice List). Estimated Volume (Text). Have you purchased before? (List). How did you hear about us? (List) .Source Channel (List) .Scheduled Visit Date (Date) .Salesperson Notes (Long Text) .4. SalesBot Automation (Triage) The Bot must adopt a human-like tone of voice, avoiding numbered menus during the initial triage. Initial Flow: Welcome message, 2-second delay, product presentation, and a question regarding the purchasing profile (wholesale or personal use). Conditional Logic: Keyword Recognition: Wholesale: "resell," "store," "supplier," "bulk," among others. Retail: "personal," "gift," "tourist," "travel," among others. 5. Management and Follow-up Automations Catalog Task: Create a task within 30 minutes of qualifying a wholesale lead. Follow-up: Automated tasks triggered after 48 hours of no response during critical stages. Re-engagement: Automated messages for leads stalled in the "Negotiation" or "Invited to Store" stages. Post-Sale: Message sent 2 hours after the sale, followed by a request for a review after 24 hours. Loyalty Building: Repurchase reminder (for wholesale) sent after 30 days. Data Hygiene: Move to "Lost" status after 15 days of inactivity. Schedule and Deliverables Timeline: Up to 15 days following receipt of the necessary data. Deliverables: Tested pipelines, active users, functional automations, and a documentation guide for future modifications. The bot-based screening process must be entirely keyword-driven (meaning there should be no numbered menus at the start), and you must strictly adhere to the specified custom field names to prevent any automation errors. Would it be possible for you to share some screenshots or portfolio samples of sales bots you have built that implement these types of conditional flows (e.g., Wholesale/Retail)? Once we determine that everything aligns with our needs, we can proceed with starting the project.