3 days ago
Tel Aviv-Yafo, IsraelStaff+
Responsibilities
- Scope customer workloads, define metrics, build benchmarks, and lead technical proofs of concept.
- Serve as the technical lead for customer engineering teams through architecture reviews and rollout planning.
- Tune inference engines using KV-cache optimization, speculative decoding, parallelism, quantization, and batching techniques.
- Profile and diagnose engine-level bottlenecks and deliver production fixes.
- Own customer onboarding, deployment, autoscaling, capacity planning, and SLO monitoring.
- Optimize configurations for base, custom, and fine-tuned models.
- Feed field insights into product roadmaps and contribute code to inference engines and tooling.
- Publish reference architectures, cookbooks, and technical blogs.
Requirements
- At least 5 years of experience in a technical role with strong programming and production skills.
- Experience with Kubernetes, Docker, and GPU servers.
- Ability to lead technical discussions with customer engineers and leadership.
- Hands-on experience with vLLM, SGLang, TensorRT-LLM, and performance profiling.
- Strong knowledge of KV-cache tuning, batching, parallelism, quantization, and related cost calculations.
- Knowledge of the open-source model landscape and hardware-model matching.
- Experience as a Forward Deployed Engineer, Solutions Engineer, or inference engineer at an AI cloud, lab, or provider.
- Hands-on experience with fine-tuning or post-training methods including LoRA, SFT, or DPO, or contributions to open-source engines.
Benefits
- Competitive benefits package supporting financial security, health, and overall well-being.
- Benefits aligned with local market standards, including pension contributions and additional work-life-balance perks.
Tech Stack
Categories
Forward Deployed
About Crusoe
Crusoe builds and operates GPU-powered data centers and an AI cloud platform for enterprises running large-scale AI training and inference. It vertically integrates energy supply with compute, using stranded and renewable power to reduce emissions and costs, and sells capacity via cloud services and managed infrastructure. Founded in 2018 and headquartered in Denver, the company is privately held.