ML Infrastructure Engineer
Graphon AI is seeking an ML Infrastructure Engineer to build the pre-model intelligence layer for the next generation of AI using a graph-based approach. The role involves developing context engines and universal infrastructure for long-context multimodal reasoning. You will work on production ML systems, training/inference pipelines, and data processing while collaborating with a lean, experienced team.
Graphon AI is building the pre-model intelligence layer for the next generation of AI. We are developing the context engine layer that solves a fundamental challenge: “What if I could just dump all my data and GPT/Claude/Gemini could smartly reason across it?” Today, this is incredibly difficult due to the limitations of LLM context windows and the fundamental squared scaling inherent to attention mechanisms. While sparse attention and RAG systems attempt to address this, they haven't proven to be robust general solutions. Pre-model intelligence is our answer. We believe a graph-based approach that analyzes relationships across data on a topological level provides the correct inductive bias for long-context multimodal reasoning. Just as CNNs provided the bias for spatial invariance in vision, and Transformers established the standard for sequential dependencies in language, our framework provides the structural foundation needed for complex, interconnected data. We believe the next wave of AI progress will come from better infrastructure around models: Better Memory & Context: Creating truly persistent, scalable state. Better Representations: Moving beyond simple embeddings to relational world models. Universal Infrastructure: A way to stop rebuilding the same context layer for every agent, dataset, and application.
Posted June 10, 2026