PascalB

AI Voice Agent Developer

Agency • Freiberufler

May 2026 - Present • 2 mos

I built an AI voice receptionist for a veterinary clinic that was fielding a high volume of calls for routine appointment bookings, prescription refill requests, and general pet-care questions, which was pulling front-desk staff away from in-clinic client care throughout the day. Using Vapi as the voice platform, I built a conversational flow trained on the clinic's service list, vaccination schedules, and common FAQs, allowing the agent to answer routine questions directly and book, reschedule, or cancel appointments by checking real-time availability through their scheduling software. For prescription refill requests, I built a flow that collected the pet's name, owner details, and medication, then automatically created a request ticket for the veterinary staff to approve, rather than the agent attempting to fulfill medical requests itself. I built in a strict escalation path for anything involving a sick or injured animal, immediately routing those calls to a live staff member rather than letting the agent attempt to triage an emergency. The agent was also configured to identify returning clients by phone number and pull up their pet's basic profile, letting the conversation feel personalized rather than starting from scratch every call. This system reduced front-desk call volume by an estimated 35%, freeing staff to focus on in-person client and patient care, while cutting the average time to book a routine appointment from several minutes on the phone down to under a minute.

Voice AI QA & Testing Specialist

TechBullion • Freiberufler

Feb 2026 - Present • 5 mos

As a Voice AI QA & Testing Specialist at VoiceOps in Toronto, Canada, I led the architecture and deployment of conversational AI systems for the tech sector. I leveraged Retell AI, Deepgram, and Twilio to engineer responsive, production-ready voice agents handling high-volume inbound and outbound interactions. My duties included designing conversation flows, integrating APIs for real-time data retrieval, and optimizing backend logic using Python for sub-second latency. A key achievement was migrating from legacy telephony to automated AI voice infrastructure, reducing operational costs by 40% and increasing lead conversions by 24%. I collaborated with cross-functional teams to build admin dashboards providing real-time analytics on call containment, transcript accuracy, and booking outcomes. By implementing advanced fallback mechanisms and context-aware memory retention, I enhanced the naturalness of the AI's conversational tone. Furthermore, I designed automated workflow orchestrations using Make.com to sync call data with Salesforce, ensuring zero data loss and strict compliance with industry regulations. I regularly monitored system performance metrics and optimized WebSocket connections to maintain peak reliability during traffic spikes, showcasing my deep expertise in NLP, telephony integration, and full-stack AI development within highly competitive markets.

High Commission SDR

Consulting • Freiberufler

Jan 2026 - Present • 6 mos

As a High Commission SDR at SDR Force AI in Dublin, Ireland, I led the architecture and deployment of conversational AI systems for the B2B sales sector. I leveraged Vapi, OpenAI, and Twilio to engineer responsive, production-ready voice agents handling high-volume inbound and outbound interactions. My duties included designing conversation flows, integrating APIs for real time data retrieval, and optimizing backend logic using Node.js for sub second latency. A key achievement was migrating from legacy telephony to automated AI voice infrastructure, reducing operational costs by 28% and increasing lead conversions by 40%. I collaborated with cross-functional teams to build admin dashboards providing real time analytics on call containment, transcript accuracy, and booking outcomes. By implementing advanced fallback mechanisms and context-aware memory retention, I enhanced the naturalness of the AI's conversational tone. Furthermore, I designed automated workflow orchestrations using n8n to sync call data with Salesforce, ensuring zero data loss and strict compliance with industry regulations. I regularly monitored system performance metrics and optimized WebSocket connections to maintain peak reliability during traffic spikes, showcasing my deep expertise in NLP, telephony integration, and full stack AI development within highly competitive markets.