Temporal Market Knowledge
Point-in-time identity, event time, observation time, lineage and versioned relationships for stocks, ETFs and portfolios.
- Temporal subgraph retrieval
- Corporate-event propagation
- Reproducible signal evidence
SDS Research & Technology Office
Applied research for the knowledge layer behind financial AI.
We study how ontology, temporal knowledge graphs, semantic standards and explainable retrieval can transform market data into governed intelligence.
Research graphACTIVE PROGRAM
Financial
AI
OntologyTemporal KGGraphRAGCyberGovernanceAI
04active research tracks
06open standards monitored
01goal: explainable financial AI
Research agenda
Where semantics can change
financial intelligence.
Each research track connects a difficult enterprise question to an implementable ontology, graph or validation pattern.
Ontology-Grounded GraphRAG
Hybrid graph and vector retrieval in which governed concepts guide extraction, traversal, ranking and explanation.
- Domain-tuned extraction
- Competency-question evaluation
- Evidence-aware responses
Cyber-Financial Reasoning
Connecting threat intelligence and technology dependencies to instruments, data products, decisions and regulatory impact.
- STIX/UCO mappings
- Business-impact paths
- Control and resilience semantics
Executable Semantic Governance
Converting data-quality, policy, access and model-risk requirements into portable graph constraints and rules.
- SHACL validation modules
- Rule provenance
- Agent action guardrails
Technical paper library
Architecture translated into
reusable engineering knowledge.
SDS / TECHNICAL PAPER 01ONTOLOGY
AS THE
AI BRAINFinancial Intelligence Architecture
AS THE
AI BRAINFinancial Intelligence Architecture
Financial Ontology · 12 min
Ontology as the Semantic Brain for Financial AI
A governed context architecture for models, GraphRAG and enterprise agents.
Read paper →SDS / RESEARCH NOTE 02TEMPORAL
MARKET
GRAPHSStocks · ETFs · Events · Signals
MARKET
GRAPHSStocks · ETFs · Events · Signals
Market Intelligence · 9 min
Temporal Market Knowledge Graphs
Preserving point-in-time truth and evidence for explainable signals.
Read research →SDS / STANDARDS WATCH 03SHACL 1.2
RULES &
GUARDRAILSEmerging Semantic Standard
RULES &
GUARDRAILSEmerging Semantic Standard
Standards Watch · 7 min
From Graph Validation to Executable Rules
What emerging SHACL 1.2 work could mean for governed AI platforms.
Read advancement →Technology advancement radar
What we adopt, trial
and watch.
Our radar separates production-ready foundations from promising capabilities that require controlled evaluation.
ADOPT
Production foundations
- RDF / OWLFormal, interoperable enterprise semantics
- SHACL 1.0Executable graph quality and constraints
- FIBOShared financial-industry concepts
- PROV-OLineage and evidence modeling
TRIAL
Controlled implementation
- Ontology-guided GraphRAGGraph + vector enterprise retrieval
- Temporal graph featuresTime-aware signal and risk context
- Semantic agent policiesGraph-constrained tool use
- FINOS CDM alignmentExecutable transaction lifecycles
WATCH
Emerging standards
- SHACL 1.2 CoreNext-generation shape semantics
- SHACL 1.2 RulesRDF-native inferencing and generation
- UCO evolutionCyber investigation model development
- Agent interoperabilityPortable semantic context services
Standards observatory
Research grounded in
open, authoritative foundations.
Latest from the research desk
Technical thinking for
builders and decision-makers.
Research partnership
Bring us a financial question that needs better meaning.
We turn the question into competency questions, a semantic model, a graph experiment and measurable evaluation criteria.