The binding constraint on an enterprise agent isn’t the model’s intelligence — it’s the context window, and the context window is denominated in tokens. This piece traces the causal chain most knowledge strategies ignore: how text becomes tokens, why the key-value cache makes those tokens the real driver of cost, latency, and memory, and why the format you choose to represent enterprise knowledge is therefore both an efficiency decision and a reasoning-quality one. It’s the argument behind SIGN (Sigil Intelligence Graph Notation), Career Highways’ open specification, measured against a full ~200-document canon — where encoding in SIGN rather than markdown cut the corpus by roughly a quarter, with the deepest savings concentrated in the most governed, structure-heavy content.
