Own DigitalOcean’s enterprise AI-native architecture, including agent orchestration, shared AI platform services, tool and capability layers, enterprise-system integrations, identity, authorization, governance, observability, evaluation, and cost controls. Build production prototypes and reference implementations, write code, establish technical standards, mentor distributed engineers, and re-engineer business workflows with functional teams. Partner across platform engineering, security, identity, data, and business technology to deliver governed, scalable agentic systems.

Lead architecture and low-level optimizations for LLM inference to maximize throughput and minimize latency. Profile and optimize GPU kernels, implement quantization and parallelization strategies (FP8/INT8/FP4), tune libraries (AITER, Triton, CUDA, ROCm), and mentor engineers while translating hardware limits into shippable platform features.

Lead design and implementation of high-scale, multi-tenant AI inference data plane services. Optimize distributed LLM serving (prefill/decode disaggregation, KV-cache routing, autoscaling, flow control), contribute upstream to open-source inference projects, mentor engineers, and operate resilient, observable production services running on Kubernetes and specialized AI hardware.