video1.0<iframe src="https://www.loom.com/embed/a08fb75e85da402db17a02e13dbf16b5" frameborder="0" width="1920" height="1440" webkitallowfullscreen mozallowfullscreen allowfullscreen></iframe>14401920Loomhttps://www.loom.com14401920https://cdn.loom.com/sessions/thumbnails/a08fb75e85da402db17a02e13dbf16b5-abfb2a38e2ba8b65.gif157.647This Loom explains how QGEN helps generate deterministic yield portfolio allocations instead of using an LLM to choose protocols. It argues that LLM-driven portfolio selection can hallucinate protocols and is non-deterministic, producing different answers each time. Using a house buyer example with $100,000 capital, a 12-month horizon, and the need for real liquidity by month 10, it shows QGEN translating a sentence into structured queries with hard filters and lens weights, then returning eight protocol options with metrics like APY, expected drawdown, audit count, and a real diversification number. It also demonstrates “vector native” viewing the same portfolio through six lenses to reveal similarity, risk clustering, correlation, and hidden concentration risk, such as how yield sources plug into downstream.