Fractional Voice AI & Conversational Automation Architect - Part-time

Xenon7 is looking for a Fractional Voice AI & Conversational Automation Architect to design, deploy and advise on production-grade voice AI agents inside enterprise customer-support environments. This is a hands-on senior role with an advisory edge. You will own the technical direction of real-time voice agent deployments — IVR, speech-to-text, text-to-speech, LLM orchestration — while also bringing senior credibility to client conversations with CIOs, IT leaders and CX/support leadership. This is not a full-time delivery role. It is a part-time engagement with flexible hours, structured around activation sprints and ongoing technical oversight. Responsibilities Architect and deploy production voice AI agents: real-time IVR, STT, TTS, and low-latency turn-taking. Design LLM orchestration and reasoning that works across both inbound voice calls and form-based support tickets. Integrate with Talkdesk or comparable CCaaS platforms, Zendesk, and backend / order-management APIs. Design intelligent routing, customer-type identification, and human escalation paths. Own the quality bar: latency budgets, voice naturalness, containment and deflection rates, plus the eval and QA approach before and after go-live. Scope activation sprints and provide oversight on delivery approach, technical feasibility and adoption planning. Review client proposals and recommendations to ensure they are realistic, enterprise-ready and aligned with platform capabilities. Join sales and discovery conversations with senior enterprise stakeholders when deep technical credibility is needed. Advise on governance, security, compliance and change management for large-scale AI deployments. Optionally contribute to thought leadership: executive insights, Microsoft or AI community events, speaking opportunities. Requirements Senior, hands-on architect/engineer who has deployed production-grade voice AI agents — shipped systems, not prototypes. Strong experience with real-time IVR, speech-to-text, text-to-speech and low-latency voice interaction design and tuning. Proven LLM orchestration and agentic workflow design across voice and text channels. Integration experience with Talkdesk or a similar CCaaS platform, Zendesk, and backend APIs. Ability to design intelligent routing, human escalation and customer-type identification. Cloud: experience deploying models in Microsoft Azure (Azure OpenAI, Azure AI Foundry) is a strong plus. If no Azure background, AWS + Bedrock experience is mandatory. Experience in high-volume, seasonal customer-support environments. Must be able to demonstrate a comparable working solution — a live demo, walkthrough or a documented past deployment. Strong architectural judgment on what is scalable, adoptable and realistic in enterprise environments. Ability to advise senior stakeholders: CIOs, IT leaders, digital workplace and enterprise transformation teams. Fluent English communication. Bonus Depth in the Microsoft AI ecosystem: Microsoft 365 Copilot, Copilot Studio, Semantic Kernel, Power Platform, Teams, SharePoint. Former Microsoft employee, Microsoft MVP designation, or experience within a Microsoft SI / partner ecosystem. Experience with enterprise Copilot adoption, licensing activation, usage expansion or business-value realization programs. Advanced academic background (Master's or PhD) in Computer Science, Information Systems, Organizational Behavior or a related technology-focused field.