Inference Engineer
Make intelligence fast, efficient and dependable.
Read job description
Future Present Labs / Join the mission
Inference Engineer
Make intelligence fast, efficient and dependable.
Hiring plan · Not a confirmed vacancy. Location, employment terms and compensation will accompany an approved opening.
Your responsibilities
Own model-serving benchmarks, batching, KV-cache behavior, quantization and GPU memory use. Profile kernels and serving paths; improve latency, throughput and cost without hiding quality regressions.
Experience and evidence
Experience profiling GPU workloads and reasoning about numerical precision, memory bandwidth and concurrency. Bring a measured optimization, reproducible benchmark or kernel you can explain.
What success looks like
Reproducible quality and performance results, reliable serving under load, and lower resource cost per useful result.
How we work
Integrity in the results you report. Persistence through difficult problems. Efficiency with compute, materials and time. Measure, tune and finish beyond the first pass—and leave the work easier for others to build on.
Start a conversation
Share relevant work, your contribution and what you learned. Tell us why this mission matters to you. Practical evidence matters more than a particular degree or title.
Introduce yourself →