About Us
We are a stealth startup building the AI platform for luxury hotels: a live digital twin of the whole hotel, with an AI agent personalising every room. We work alongside technology leaders in AI and with luxury hospitality brands.
Our working prototype is a live twin of a luxury hotel room, rendered in real time and driven by a GPU-based building simulation. Now we take it from one room to a whole hotel.
About the role
Reports to: Head of Engineering
Every room in our hotels has its own AI agent. It knows the guest's preferences, controls the room (climate, lighting, curtains, TV), talks to the guest, and works with staff and hotel systems behind the scenes. As Lead Engineer, Agentic Systems, you will design and build that agent platform from the first line of code, and lead the agentic side of the team.
Responsibilities
- Design the agent architecture: a per-room agent, a hotel-level orchestrator, and specialist agents for housekeeping, maintenance, concierge and energy.
- Build tool use against real room controls and hotel systems, through the digital twin's live state.
- Own memory and personalisation: guest preferences that carry across stays, with privacy and consent built in.
- Build evaluation: scenario suites, simulated guests, regression tests and a trace of every agent action.
- Build guardrails: permissions, human sign-off for staff-facing actions, and safe fallbacks when a model or the network fails.
- Ship low-latency voice and chat interfaces for guests and staff.
- Choose models and serving: frontier APIs and open models on NVIDIA GPUs (NIM, TensorRT-LLM), balancing cost, latency and quality.
Qualifications
- 3+ years of production software engineering, with at least a year shipping LLM-based products to real users.
- Hands-on with agent and tool-use patterns: function calling, MCP, LangGraph or similar, or your own orchestration.
- Strong Python or TypeScript, and solid backend fundamentals (APIs, queues, databases, event-driven systems).
- Experience building evals and debugging non-deterministic systems.
- Product sense: you think about the guest's experience, not only the model.
- Real-time voice agents (speech-to-speech, telephony) are a big advantage.
- Fine-tuning or distilling open-weight models, NVIDIA NIM, Riva or NeMo, smart-home or IoT control, or hospitality tech experience is a strong plus.
The details
- Day rate: £575 to £800 per day.
- Location: in person, five days a week, at our office in central London.
- Contract: 6 months, full-time, as a contractor or on a fixed-term contract. Strong performers can convert to a permanent role with equity. Contracts are with Ellie Technologies Ltd.
- IR35: Ellie Technologies Ltd is a small company, so contractors working through their own limited company assess their own IR35 status. A fixed-term employment contract on payroll is also available.
- Start: as soon as possible. We are interviewing now.
- Right to work: you need the existing right to work in the UK. We cannot sponsor visas for these roles.
- Process: a 20-minute intro call, a 90-minute technical session, then an offer. We aim to decide within a week of your application.
We are an equal opportunity employer. We welcome applicants from every background.