Du’An Lightfoot is a senior AI engineer at Akamai and the creator of LabEveryday, an educational platform for engineers developing practical technology skills. His career spans enterprise networking, cloud infrastructure, autonomous agents, and the operational demands of running language models on dedicated hardware.
An Air Force veteran, Lightfoot worked in technical support, systems administration, and enterprise networking, including at Cerner, before joining Cisco as a DevNet content developer engineer. There, he helped infrastructure engineers adopt Python, APIs, Ansible, and network automation through technical writing and developer education.
He created LabEveryday while pursuing networking certifications, motivated partly by the scarcity of Black educators in the technical communities where he was learning. He later joined AWS as a senior cloud networking developer advocate, expanding into generative AI and agent development.
At AI Engineer World’s Fair 2025, Lightfoot co-led a workshop on browser agents and the Model Context Protocol with Banjo Obayomi. His contribution emphasized agents that plan, use tools, evaluate outcomes, and adjust their behavior, while reserving deterministic automation for simpler tasks.
Lightfoot subsequently moved from AWS to Akamai, where he built reusable Kubernetes workshop infrastructure for GPU-backed inference. His work now focuses on self-hosted AI inference, vLLM, dedicated accelerators, workload isolation, and production reliability.
- Inference needs infrastructure-aware engineering. Lightfoot treats GPU memory bandwidth, key-value cache pressure, batching, tail latency, admission control, and external queues as first-class production concerns. An available endpoint can still fail users when concurrent demand overwhelms the system, as he details in his analysis of inference failure modes.
- Dedicated GPU inference gives agent builders operational control. His vLLM-on-LKE deployment project creates an OpenAI-compatible model endpoint on Linode Kubernetes Engine, with monitoring, load testing, security guidance, and agent integration. The approach addresses privacy, compliance, latency, rate limits, and unpredictable token costs.
- Network MCP makes infrastructure actionable for agents. His open-source network diagnostics server equips agents to investigate connectivity, routing, DNS, and packet behavior, applying his networking background to tool-enabled AI systems.
- Agent evaluation and human oversight determine whether automation is dependable. Lightfoot stresses continuous evaluation, guardrails, verification, and human intervention when browser sessions, authentication, CAPTCHA challenges, or ambiguous tasks exceed an agent’s practical capabilities.
