Senior Voice AI Engineer - Contractor position

We're a small UK AI automation consultancy hiring a senior voice agent engineer to lead our voice AI capability. First major engagement is a clinical-grade voice agent for a UK private healthcare provider. Read the role below; the application takes around 20 minutes via Typeform and we respond to every applicant within 5 working days.

• Location. UK (London or hybrid). Some on-site discovery days with customers required.
• Engagement. Permanent or long-term contract considered.
• Comp. Senior band, market-competitive. [TBD: fill in before posting]
• Reports to. Lydia (CTO) on tech; Adil (Automation PM) on project delivery.

About ivee

ivee is a UK AI automation consultancy delivering embedded AI builds for B2B customers. We replace manual operational processes with production-grade AI inside the customer's existing tech stack. We work on fixed fees with milestone payments, and we hand over working systems that the customer's team can run.

We're scaling our voice AI capability. Our first major voice engagement is a UK private healthcare provider (10 clinics, around 180 staff, 2,500 inbound calls per week) deploying a clinical-grade voice agent that integrates with their EMR and PBX. That's the engagement you'd lead.

About the role

This is a build-and-own role. You won't hand off after design; you'll lead the technical architecture, vendor selection, build and go-live for production voice agents. You start with the healthcare engagement above and expand into the next couple of customers as the capability matures.

You'll work directly with customers. Running on-site discovery workshops, defending architecture choices to C-level stakeholders, owning the technical conversation through to go-live. You're not behind a PM; you're the technical lead.

What you'll do

• Architect voice agent deployments end to end. Telephony integration, orchestrator selection, prompt engineering, tool integration into customer LOB systems, monitoring, post-launch hypercare.
• Lead technical due diligence on customer environments: telephony stack, EMR/CRM APIs, compliance overlay, latency budgets.
• Stand behind vendor selection. Vapi against Retell against Synthflow against Pipecat against LiveKit against Bland against ElevenLabs Conversational AI. You'll be expected to defend your choice with rationale grounded in the customer's specific brief.
• Design healthcare-grade safety guardrails. Identity verification, vulnerable-caller handling, ICO-defensible call recording posture, clinical safety scope (keeping the agent out of MHRA Software-as-Medical-Device territory).
• Lead two-day on-site discovery workshops with customer stakeholders (typically C-level plus operational leads). Output is a scoping document the customer can stand behind.
• Mentor the wider ivee engineering team on voice AI patterns so the capability scales with the company.
• Contribute to ivee's productised voice AI offering as we move from custom-build engagements to a more repeatable pattern.

Must-have

• Three-plus years building production voice AI systems. Not POCs, not chatbots, not "voice-enabled FAQ." Systems handling real customer calls at production volume for at least three months.
• Hands-on production experience with at least two of: Vapi, Retell AI, ElevenLabs Conversational AI, Pipecat, Synthflow, LiveKit Agents, Bland AI.
• Deep SIP and telephony knowledge. SIP trunks, REFER handling, codec negotiation, NAT traversal, Twilio Elastic SIP or Telnyx. Comfortable reading SIP traces.
• Production tool-calling integration into customer LOB systems. CRM, EMR, scheduling, finance systems. Not just FAQ retrieval.
• TTS and STT engineering. ElevenLabs, Deepgram, Cartesia, PlayHT. Voice quality tuning, barge-in handling, latency optimisation.
• LLM application engineering. Prompt engineering for production, latency budgeting, parallel tool calls, retrieval, eval suites. You understand the difference between "works in a demo" and "works at 9am on a Monday."
• Track record shipping into regulated environments. Healthcare, finance or legal. Comfortable treating compliance constraints as design inputs rather than blockers.

Strong nice-to-have

• 3CX experience. Specifically v20, attended-transfer behaviour, CRM integration template, call flow designer. If you've seen 3CX's warm-transfer bugs in production and worked around them, lead with that.
• Meddbase or a comparable UK private-healthcare EMR (DrDoctor, Healthcode-integrated systems).
• UK GDPR fluency for AI processing. DPIA contribution, ICO AI guidance, identity-verification design under ICO scrutiny.
• CQC, DCB0129 or DSPT awareness, even at a "have-read-the-docs" level.
• Healthcode or VEDA familiarity for the financial and claims side.
• Open-source contributions in the voice AI space (Pipecat, LiveKit Agents, Vapi tools, etc.).
• ElevenLabs Professional Voice Cloning experience, including the consent and onboarding workflow.

Soft requirements

• Comfortable in customer-facing technical discussions, not just engineering Slack.
• Can lead a two-day on-site discovery workshop with C-level stakeholders without ivee hand-holding.
• Writes clean architecture documents and technical specs. Collaborative output, not "trust me" code.
• Pragmatic about platform choice. Opinion with reasoning, not religion.
• Comfortable with consultancy engagement model (fixed scope, customer accountability, milestone payments).

What we offer

• Lead role on a fast-growing automation engagement portfolio.
• Senior comp band.
• Hybrid working with London flexibility.
• Direct customer engagement; you're not behind a PM.
• Genuine ownership of a capability we're building, not maintaining.

Click apply to fill in a short form and we'll get back to you ASAP!