Build AI Sales Coach for RingCentral Text Conversations (NOT a Chatbot)

Overview I own an automotive restyling business specializing in window tint, paint protection film (PPF), ceramic coatings, detailing, and hail repair. We use RingCentral for inbound and outbound SMS communication with leads and customers. I am NOT looking for a chatbot. I am NOT looking for an AI auto-responder. I am looking for an AI Sales Coach that reviews SMS conversations the same way an experienced sales manager would and advises my team on the best next step to maximize the probability of closing the sale. The primary purpose of the system is sales analysis and coaching, not customer communication. Project Goal When a new text message arrives in RingCentral, the system should: 1. Retrieve the conversation history. 2. Analyze the full conversation thread. 3. Determine where the lead is in the sales process. 4. Identify objections and buying signals. 5. Recommend the best next action for the salesperson. 6. Deliver the analysis to Slack or a similar platform. Required Functionality RingCentral Integration * Monitor inbound and outbound SMS conversations. * Trigger analysis when new messages arrive. * Retrieve conversation history automatically. * Maintain conversation history by lead. Conversation Analysis The system should analyze the entire conversation thread, not just the latest message. The AI should identify: * Service type * Lead quality * Buying intent * Lead stage * Objections * Customer engagement level * Whether the lead is progressing or stalling * Whether a phone call is more appropriate than texting Sales Coaching Output For each conversation provide: Lead Score (1-100) Close Probability (%) Service Type * Window Tint * PPF * Ceramic Coating * Hail Repair * Detailing * Unknown Lead Stage * New Inquiry * Information Gathering * Quote Sent * Objection Handling * Follow-Up Needed * Ready To Schedule * Existing Customer * Lost Lead Primary Objection Examples: * Price * Timing * Needs Spouse Approval * Comparing Competitors * Distance * Insurance Questions * No Objection Identified Recommended Next Action Examples: * Send Estimate * Follow Up * Ask For Deposit * Ask For Appointment * Call Customer * Send Social Proof * Stop Educating And Ask For The Sale * Move To Long-Term Nurture Reasoning The AI should explain WHY it is recommending the action. Example: “The customer has already received pricing, warranty information, and product education. No additional objections have been raised. The next step should be scheduling rather than additional education.” Suggested Reply A suggested reply may be generated, but this is secondary. The primary purpose of the system is sales coaching and next-step recommendations. Training Data I have historical SMS conversations that are currently being organized and categorized. The system should be designed so that: * Historical conversations can be imported. * Future conversations can be stored. * Business knowledge can be updated. * Pricing and sales processes can be updated. * Additional integrations can be added later. Deliverables & Milestones Milestone 1 – Architecture & Planning Deliverables: * Architecture diagram * Written explanation of system workflow * Explanation of how conversation history will be stored * Explanation of how AI recommendations will be generated * Project timeline No development should begin until architecture is approved. Milestone 2 – Functional MVP Deliverables: * RingCentral integration * Conversation retrieval * Conversation storage * OpenAI integration * Sales coaching output * Slack notification delivery * Loom/video walkthrough Success Criteria: When a new text message is received, the system automatically analyzes the conversation and returns: * Lead Score * Close Probability * Service Type * Lead Stage * Primary Objection * Recommended Next Action * Reasoning * Suggested Reply Milestone 3 – Refinement Deliverables: * Improved scoring and recommendations * Ability to upload historical conversations * Documentation * Final Loom/video walkthrough Important I am not a software developer. Deliverables should be presented in a way that a non-technical business owner can understand and verify. Screenshots, diagrams, Loom videos, and visual demonstrations are strongly preferred. If your proposal focuses primarily on building a chatbot or AI auto-responder, you have misunderstood the project. I am trying to build an AI sales manager, not an AI customer service representative. Please include in your proposal: * Similar projects completed * Recommended architecture * Estimated timeline * Estimated cost * Whether you would store conversation history in a database and why * Any recommendations you would make to improve this concept