Jarrod Smith

I am a Principal AI/ML engineer with over 11 years designing, architecting, and deploying large-scale AI solutions across enterprise, cloud, and embedded ecosystems. I specialize in transformer-based models, generative AI, retrieval-augmented generation (RAG), and multimodal machine learning.

At my current role I led end-to-end LLM-powered recommendation engines and MLOps infrastructures, leveraging PaLM, T5, Apache Beam, and Vertex AI, which delivered a 28%+ uplift in personalization KPIs. I built scalable low-latency RAG pipelines and production inference architectures with Kubernetes, TensorFlow Serving, and Prometheus-based observability to ensure availability and automated rollback.

In prior roles I designed GPT-powered assistants embedded in CRM systems, applied multilingual fine-tuning to improve triage accuracy by 31%, and orchestrated CI/CD model deployment pipelines that reduced deployment timelines by 44%. I have benchmarked and fine-tuned large language models, integrated vector DBs and search stacks, and produced interactive dashboards to guide stakeholders on model trade-offs and KPIs.

I mentor engineers in advanced LLM deployment, prompt engineering, PEFT (LoRA/QLoRA), RLHF, and acceleration tools, and I enjoy leading cross-functional Agile squads to deliver measurable business impact. I seek to build production-grade AI systems with strong MLOps, responsible-AI practices, and clear operational metrics.