Voice AI. The kind that picks up.
Moonlabs is the operator-led AI Academy in Derby. We run three live companies — Homemove, home.co.uk and homedata.co.uk — and we teach twelve students per cohort to ship a real AI product, sell it to a real customer, and raise on it. Three pillars: Coding, Commercials, Investment. Twelve weeks. £6,000.
Moonlabs is what we are. Two operators — James Freestone and Louis O’Connell-Bristow — who run Homemove, home.co.uk and homedata.co.uk. Homemove handles inbound phone enquiries at volume — voice AI is a real production surface for us, not a side experiment. The latency budgets and eval patterns on this page are the ones we tuned against real callers.
The Academy is what we do. A twelve-week, in-person, twelve-student cohort in Derby. You build a real AI product. You sign a paid pilot on it. You write a deck and a financial model. You leave with a deployed system, a paying customer reference and a live investor pipeline. Coding, Commercials, Investment — the three pillars taught in equal weight every week.
Why this page exists. Voice AI was the AI category most underrated in 2024 and most over-promised in 2025. Real-time speech-to-speech models are now genuinely good, latency is within the range a phone call tolerates, and the economics work for the right use cases. What is still rare is the engineering discipline to ship one that actually answers the phone reliably. You leave the Academy with a voice agent taking real calls from real users — or as the founder of a voice-AI product targeting one of the loudest live thesis areas in venture.
Coding · voice AI at production fluency
OpenAI Realtime API, ElevenLabs, Cartesia, Deepgram, AssemblyAI. Twilio / Vonage / SIP telephony bridges, real telephone numbers, real callers. Sub-second latency budgets, streaming, interruption handling, function calling mid-utterance, transcript-based evals, hand-off to humans without dropping context. A deployed voice agent answering real calls by week twelve.
Commercials · the AI receptionist economics
A reliable AI receptionist replacing one human FTE saves an SMB £25-40k/year — one of the highest-ROI AI offers a small team can sell. Pricing per inbound minute or per seat replaced, the discovery call, a one-page pilot agreement, the first paying customer. A paid pilot by week six — for many graduates, a productised “your front desk, AI-answered” offer.
Investment · raising on voice-AI
ElevenLabs ($3bn), Sesame, Cartesia ($91m from a16z), Vapi, Bland AI ($16m), Retell AI, PolyAI ($500m+), Hume, Wispr Flow, Sindarin, ConversaSpace, Decagon ($1.5bn on AI customer support) — voice-AI is one of the densest funded categories of the cycle. Cap table, ten-slide deck, financial model. A live investor pipeline by demo day.
Common questions.
Which voice providers do you cover?
OpenAI Realtime API as the current default for speech-to-speech, ElevenLabs and Cartesia for high-quality synthesis, Deepgram and AssemblyAI for transcription. The course is deliberately provider-agnostic — the patterns transfer cleanly between stacks.
Do I need to know telephony to apply?
No. We will get you fluent in Twilio (or equivalent) inside the first fortnight. Telephony has its own culture and gotchas but the learning curve is shorter than people fear.
Will I build a voice clone of myself?
Lightly — voice cloning is covered where it earns its place. Most of the course is about agents that handle real calls reliably, which is a different (and harder) problem than cloning a single voice well.
What kinds of voice agents do students build?
Past projects include receptionists for SMBs, outbound qualification agents for B2B sales teams, support triage for ed-tech, and inbound booking agents for service businesses. The pattern is “a call a buyer would pay to have answered”.
How does this fit with the wider agents and LLM curriculum?
Voice is one delivery surface for the same underlying agent architecture. The agents course covers the architecture; this page covers the voice-specific application of it.
More Academy entry points.
The Academy is one course with many doors. Each of these pages is a different entry point into the same twelve weeks.
Build it. Sell it. Raise on it. In twelve weeks.
Tell us what voice agent you would build and who would pay for it. James and Louis read every application personally and reply inside the week.
© 2026 Moonlabs Incubator. All rights reserved.