Research Engineer, Language Modelling
Zurich, CH · hybrid
About ORIQX
Compute is fragmenting. For a decade the answer was one CPU and one class of GPU. It is now many: further GPU classes, gate quantum processors, annealers, analog and thermodynamic samplers, wafer-scale parts and custom accelerators, each an order of magnitude better or worse depending on the problem, and each with a different cost, energy and error profile. The decision of what to run where is currently made once, by hand, and is wrong as soon as prices, hardware or problem size change.
ORIQX is three things: a model, an engine and hardware. The ORIQX compute model is being built, starting now: a model that works in the vocabulary of a domain and states what to compute, never where. The ORIQX compute engine is live today: before anything executes, it predicts what a computation will cost in time, money, accuracy and energy on every class of chip we model, runs each part where it wins, and returns a report of predicted against actual. ORIQX hardware is being simulated now, and its blueprint follows in twelve months: the design of a heterogeneous rack built from other vendors' silicon, sized to a real portfolio of workloads.
The engine sits beneath the orchestrators rather than among them, and all three are meant for every vendor's silicon rather than one.
About the role
This seat owns how a language model is in fact served underneath an agent: batching, cache behaviour, where a decode step runs, and what the memory wall does to the cost of a token on each class of chip the company models.
Decode re-reads the full weight set per token, which makes language-model serving a memory-bandwidth problem before it is a compute problem, and makes the choice of chip class for it a different decision from the one the rest of the platform makes. This seat owns that decision: the batching and cache policy, the split between prefill and decode, which class of silicon each half is offered to, and the measurement that says whether the split was right. It sits beneath the agent loop rather than inside it, and beneath the orchestrators rather than among them.
Responsibilities
- The surface has a measurement before it has an optimisation. What the loop costs today, per completed task and per class of chip, recorded well enough that a change can be shown to have helped.
- One routing decision inside the loop is made by the platform rather than by hand, and the decision is shown to be better than the fixed choice it replaced on a workload somebody outside the team can re-run.
- The failure behaviour is specified and tested: what a cancelled step does, what a retried tool call costs, and what the accounting says in both cases. A loop that is only correct on the happy path is not a runtime.
You may be a good fit if you
- The candidate has built or substantially owned a runtime, scheduler or serving layer that other people's code ran on, and can describe a failure mode it had in production.
- The candidate measures before optimising, and can describe a change they reverted because the measurement did not support it.
- The candidate is comfortable that the interesting number here is cost per completed task, not tokens per second, and can explain why those come apart.
- Coding agents are used as a matter of course and what they produce is verified; that is how this company works.
Strong candidates may also
- Experience with inference serving at a scale where batching and cache policy mattered to the bill.
- Having owned the accounting side of a system as well as its performance, so a cost figure could be defended to whoever paid it.
Logistics
- Education
- A degree in computer science, electrical engineering, physics or mathematics is one way to be ready for this work. Equivalent experience is another, and the two are weighted equally. The assessment is what the candidate has built.
- Location
- Based in Zurich, or hybrid. Relocation support.
Apply even if you do not meet every line above. Few strong candidates meet all of them, and the ones who exclude themselves early are disproportionately the people we would want to have met.
Come work with us
- What you can do
- We care about what you can do, not where you learned to do it.
- Compensation
- Every seat has a real band, and it is given on the first call: competitive for a company at this stage, and not matched to a frontier laboratory's cash.
- Equity
- Equity is part of every offer. The instrument, the vesting and the Swiss tax treatment are settled in the offer letter, with the candidate. No range is published, because none is agreed for a seat until then.
- Where
- Based in Zurich, or hybrid. Relocation support.