Senior AI Automation Engineer - 4 Production Systems I need ONE technical lead to deliver four confirmed AI automation projects for Quebec clients. Total value: $93k. These are production systems, not prototypes. CRITICAL: One project is a real-time voice AI phone system requiring <2s latency. No voice AI experience = don't apply. THE PROJECTS Project 1: AI Phone Receptionist (8 weeks) Fuel company needs bilingual (FR/EN) voice AI handling calls 24/7. Provides pricing, books deliveries, checks CRM, transfers to humans when needed. <2 second response time mandatory. Tech: Twilio Voice, Deepgram/AssemblyAI, ElevenLabs, GPT-4 function calling, CRM integration, production reliability. Project 2: Collections Automation (7 weeks) Same client. Auto-send payment reminders from QuickBooks via SMS/email. Personalized escalation, incentive offers, CRM sync, daily reports. Tech: QuickBooks API, SendGrid, Twilio SMS, GPT-4, scheduling logic. Project 3: RFP Processing (2 weeks) Manufacturing company. Parse 100-300 page RFPs, extract materials/specs with AI, contact suppliers, compile bids automatically. Tech: PDF parsing, GPT-4 extraction, email automation, quote comparison, document generation. Project 4: Roofing Automation (2 weeks) Auto-schedule jobs from QuickBooks estimates, order materials, adjust for weather, notify team. Tech: QuickBooks API, Google Calendar API, Weather API, notifications. YOU MUST HAVE Voice AI (Non-Negotiable): Production voice systems with real customer calls Twilio Voice or similar (WebSocket streaming) STT/TTS integration (Deepgram, ElevenLabs, etc.) Achieved <2s latency before Bilingual voice systems (French + English) Production error handling AI/LLM: GPT-4 API with function calling Prompt engineering for production Error handling Integrations: QuickBooks API (OAuth, webhooks) CRM APIs Google Calendar API Email/SMS APIs (SendGrid, Twilio) Development: Python (FastAPI/Flask) OR Node.js Async/concurrent programming RESTful APIs PostgreSQL/MongoDB PDF document parsing Cloud hosting (Railway, Render, AWS) Language: Fluent French + English (clients are Quebec-based) YOU WILL DELIVER Four working production systems - deployed, tested, running 24/7 Complete documentation - architecture, setup guides, troubleshooting All integrations configured - APIs, webhooks, OAuth flows working Knowledge transfer - video walkthroughs, 2-3hr training, 30 days support TIMELINE Duration: 1-2 months Hours: 30-40/week Start: Within 1-2 weeks Compensation: Discussed with qualified candidates APPLICATION REQUIREMENTS (ALL MANDATORY) 1. Voice AI Portfolio Link to working voice AI you built (we must be able to call/test it) Video demo Architecture explanation How you achieved low latency If no voice AI experience: Explain similar real-time systems built and your specific learning plan. 2. Answer These 3 Questions (300 words max each): Q1: Design a bilingual voice AI receptionist with <2s response time that checks CRM, books appointments, and transfers to humans. Describe your tech stack, latency strategy, error handling, monthly costs, and biggest risks. Q2: Customer owes $5,432 (45 days overdue) with history of late payment. Design the automated collections logic: message content, tone, incentives, human escalation point, response tracking. Show decision tree or pseudocode. Q3: Client reports: "Your AI has 5-second delays and cuts off mid-sentence." Walk through: debugging process, what causes each issue, fixes, prevention, monitoring. 3. Code Sample GitHub link to relevant project (voice AI preferred) OR pseudocode for WebSocket audio streaming server 4. References 2-3 similar projects completed Client testimonials (if available) 5. Availability Hours/week, start date, timezone SCREENING PROCESS Stage 1: Application review (3-5 days) Stage 2: 30-min technical interview (screen share past project, live problem-solving) Stage 3: Paid test project ($500, 5-7 days) - build mini voice AI with <3s latency AUTOMATIC REJECTION ❌ No voice AI portfolio ❌ "I can learn voice AI as I go" ❌ Claims 4 weeks for all projects ❌ Only knows Make.com/Zapier ❌ Copy-paste generic proposal ❌ Can't explain latency optimization ❌ Won't do paid test WHAT WE LOVE ✅ Working voice AI demo in application ✅ Realistic timeline (8+ weeks for phone system) ✅ Asks detailed questions about edge cases ✅ Suggests starting with simpler project first ✅ Shares past failures + lessons learned ✅ Proposes improvements to requirements WHY THIS IS GREAT Real $93k revenue, signed contracts Interesting technical challenges Production voice AI for portfolio Long-term maintenance potential Partnership opportunity Clear requirements WHY THIS MIGHT NOT FIT High client expectations Timeline pressure Some client hand-holding needed Working solutions > perfect code You're solo technical lead Voice AI portfolio link 3 technical question answers Code sample/GitHub References Availability No AI portfolio = not reviewed Deadline: Oct 17, 2025 Decision: Oct 20, 2025 Start: Oct 21, 2025