Description
Summary AI Voice Agent for Towing & Roadside Assistance — Phase 1 Proof of Concept We are looking for an experienced AI voice developer to build the first proof of concept for an AI phone agent for a towing and roadside assistance business. The fixed budget for this initial Phase 1 is $500. This is NOT the budget for the complete production system. Our goal is to first prove that the core voice workflow works reliably. If Phase 1 is successful, we plan to continue development in additional paid phases with the same developer. Phase 1 goal Build a working AI voice agent that we can test by calling a real phone number. The AI should answer the call, communicate naturally with the customer, understand the type of roadside service needed, collect the essential information, confirm it, and output the collected information in a structured format. For this $500 Phase 1, the AI agent should be able to: * Answer a real inbound phone call using Twilio or a comparable telephony provider. * Have a natural real-time voice conversation with reasonably low latency. * Identify the requested service, such as towing, jump start, flat tire, vehicle lockout, or fuel delivery. * Collect the customer’s name and callback phone number. * Collect the customer’s current/pickup location. * Ask for the destination when towing is required. * Collect basic vehicle information such as year, make, and model. * Ask basic follow-up questions based on the requested service. * Confirm the important information with the customer before completing the call. * Produce/store the collected request information in a structured format (JSON, webhook, database, or another simple method that can later be integrated with our dispatch system). Important: Phase 1 does NOT require a complete dispatch platform, payment system, driver management system, advanced quoting system, or full production infrastructure. Those features may be developed in future phases after the initial voice-agent concept has been successfully tested. Technology We are open to recommendations on the best stack. Experience with technologies such as Twilio, OpenAI/LLM APIs, OpenAI Realtime API, Deepgram, ElevenLabs, Python, WebSockets, speech-to-text, text-to-speech, and real-time voice applications is relevant. We are not committed to any specific vendor. We want the developer to recommend a reliable and cost-effective architecture. Phase 1 deliverable A working proof of concept that allows us to call a real phone number and complete a realistic towing/roadside assistance conversation with the AI from beginning to end. We should also receive the source code and basic setup/documentation necessary to test and continue development of the project. Future phases may include: * Integration with our dispatch workflow/software * Automatic price/quote calculation * SMS communication * Payment processing * Driver assignment and communication * Customer notifications * More advanced AI decision-making * Additional automation and integrations When applying, please answer these questions: 1. Have you built a real-time AI voice/phone agent before? Please provide an example or demo if available. 2. What technology stack would you recommend for this Phase 1 and why? 3. Can you complete the described Phase 1 proof of concept within the $500 fixed budget? 4. What is your estimated delivery time? 5. What would you need from us to begin? Please start your proposal with “TOWING AI” so we know you have read the complete job description. We are looking for someone who can potentially continue with us through the next development phases if the initial proof of concept is successful.