DecisionSynth Full Decision Corpus
The complete commercial decision corpus: 1,979 deterministic advisor-decision episodes with 4,913 QA evaluation tasks and answer keys over 1,752 fresh synthetic households spanning all 71 client archetypes. Every episode records trigger, options, recommendation, override, outcome, and cited IRS/Treasury rule keys — ground truth known by construction, version-stamped in _meta, byte-identical on regeneration. Disjoint by construction from the free DecisionSynth Bench dev set and the private held-out set.
Use Cases
Technical Highlights
Sample Schema Fields
{
"decision_type": <value>,
"trigger.event_type": <value>,
"options[].label": <value>,
"recommendation": <value>,
"resolution.override_reason": <value>,
"resolution.outcome": <value>,
"semantic_refs[]": <value>,
"_meta.domain_versions": <value>
}Related Wealth Data Sets
137 advisor-decision episodes over 277 synthetic households — the same 6 client archetypes as the Reg BI Suitability Audit Pack. Each episode records what triggered the decision, the options weighed, the recommendation, any override and its reason, the outcome, and the IRS/Treasury figures that governed the choice (cited by Rule Sets fact key). Deterministically generated, so every QA task ships with a ground-truth answer key — evaluate an agent-memory pipeline against answers known by construction, not labeled after the fact.
282 advisor-decision episodes over 230 synthetic households — the same 10 client archetypes as the Tax-Loss Harvesting Simulator. Each episode records what triggered the decision, the options weighed, the recommendation, any override and its reason, the outcome, and the IRS/Treasury figures that governed the choice (cited by Rule Sets fact key). Deterministically generated, so every QA task ships with a ground-truth answer key — evaluate an agent-memory pipeline against answers known by construction, not labeled after the fact.
254 advisor-decision episodes over 530 synthetic households — the same 10 client archetypes as the Retirement Income Sequencing Pack. Each episode records what triggered the decision, the options weighed, the recommendation, any override and its reason, the outcome, and the IRS/Treasury figures that governed the choice (cited by Rule Sets fact key). Deterministically generated, so every QA task ships with a ground-truth answer key — evaluate an agent-memory pipeline against answers known by construction, not labeled after the fact.
Purchases are for internal use only. Redistribution or resale of data is prohibited under the WealthSchema Data License.
View data license →