The graph is taking shape. Two months of ingested data, and the lattice structure is emerging naturally from the decision extraction process. The patterns are real.
The Node Structure
Each decision node contains: the decision itself (what was chosen), the signals that triggered it, the constraints that bounded it, the alternatives that were considered and rejected, the stakeholders involved, the risk assessment (explicit or inferred), and the outcome where known.
Outcomes are critical but tricky. For some decisions, the outcome is clear and documented. For others, especially long-range strategic decisions, the outcome is still unfolding or was never explicitly captured. Tessera handles this with confidence gradients rather than binary success/failure labels.
The Edge Structure
Edges connect decisions that share structural properties. Two decisions are linked if they share: similar constraint profiles, similar risk/reward asymmetries, similar stakeholder dynamics, similar domain collision patterns, or similar temporal patterns (time pressure, deliberation period, escalation timing).
The edges are weighted. Strong structural similarity creates strong edges. Weak or partial similarity creates weak edges. The graph self-organizes into clusters of structurally similar decisions, and those clusters span domains and decades.
What the Graph Reveals
I can already see patterns I was not consciously aware of. There are clusters of decisions where I consistently favor speed over thoroughness, and they share a specific risk profile: low irreversibility and high information decay. There are other clusters where I consistently slow down despite pressure, and they share a different profile: high irreversibility and asymmetric downside. The graph is making my own meta-heuristics visible to me for the first time.