Meramia already ships agents and retrieval onto client platforms. This role finds the structural waste in that path and removes it: end-to-end latency, memory bandwidth, GPU occupancy, and the bill for a token that did not need to be spent.
You will measure first. A change that cannot be scored against p99, throughput, or serving cost does not land. Accuracy is a constraint, not a thing we trade away to look fast.
What you will do
- Walk the critical path from an agentic loop to the inference pipeline and the container that holds it, and name the hop that is actually expensive.
- Build the telemetry and profiling the rest of the team can run: latency, memory bandwidth, compute throughput, GPU utilisation, serving cost on the clusters that already exist.
- Land architecture changes with the people who train, serve, and productise the models — fewer wasted tokens, tighter context handling, a topology that does not multiply RPCs for sport.
- Tune serve-side workloads: hardware utilisation, context pipelines, microservice shape. The number that moves is the one an operator can feel or a finance lead can read.
What you bring
- ML inference or serving platforms and large cloud estates in the same week: Kubernetes, distributed services, GPU orchestration.
- A habit of chasing latency and memory through a high-throughput system with traces, profiles and benchmarks you can replay.
- A production codebase that was messy when you arrived, and a win you can point at after — milliseconds or dollars, written down.
- The trade: raw speed against compute cost, reliability, and the person waiting on the token stream.
Useful, not required
- CUDA kernels, Triton, vLLM, TensorRT-LLM, or a C++ / Rust runtime you extended rather than wrapped.
- Agent loops, context-window management, streaming responses — how they fail when the KV cache or the network is the real budget.
- Open-source work on ML systems, HPC libraries, or cloud-native tooling.
What to send
- A CV and the two profile links the form asks for.
- One optimisation you owned: the bottleneck, the measurement, the change, and what it did to p99 or the serving bill.
The first quarter
The first month is an audit of latency and cost on the paths we already run — inference, the network, the agent loop. The second is a single profiling and telemetry baseline the team can read: latency, throughput, GPU memory. The third is one landed cut: a named drop in p99 or serving cost on a primary workload.