Context Graph &Decision Intelligence

Engine for  AI Agents, LangGraph Workflows, Claude Code, CrewAI Teams, LlamaIndex Pipelines, AutoGen, Production AI, Enterprise AI, Any AI Framework

The Challenge

The world's oceans generate an estimated $2.5 trillion in economic value annually: fisheries, coastal protection, tourism, and carbon sequestration. Yet this natural capital remains almost entirely invisible to the financial system. The problem is not a lack of data. OBIS holds over 100 million marine species occurrence records. Copernicus satellites pass over every reef and mangrove bed multiple times daily. Hundreds of peer-reviewed papers quantify ecosystem service values with confidence intervals. The problem is that none of this translates automatically into a financial number that capital can underwrite.

How Semantica Helped

Semantica is the provenance engine inside Nereus. It ingests peer-reviewed scientific literature, extracts causal claims with full DOI attribution, and assembles them into a deterministic financial inference graph. The core mechanism is the Bridge Axiom: a conditional financial primitive derived directly from a scientific paper and permanently linked to its source. An AI agent traverses the verifiable graph path from a raw satellite or sensor reading, through one or more validated axioms, to a calculated financial exposure metric. The output is a P5/P50/P95 exposure figure with a complete, reproducible audit trail from science to finance. The system also generates TNFD-compliant LEAP disclosure packages on demand, as a byproduct of the analysis rather than a separate workflow.

The Challenge

Each AER regulatory cycle produces thousands of pages across a structured document hierarchy: Regulatory Proposals, AER Issues Papers, AER Draft Determinations, Ausgrid Responses, Expert Witness Reports, Alternative Control Services determinations, and Final Determinations. Two complete cycles, 2019–24 and 2024–29, already exist in full. They contain every objection the AER has raised, every cost category Ausgrid has defended, every accepted and rejected justification, and every negotiated outcome. The AER references this history explicitly when assessing new proposals. Ausgrid's submission teams were not.

How Semantica Helped

Semantica ingested the complete public corpus of AER and Ausgrid regulatory documents across both cycles, parsing PDFs with structure-aware chunking that preserves the regulatory document hierarchy rather than breaking at arbitrary character limits. The output is a temporal knowledge graph in Neo4j where every extracted entity is linked to its source document, part, section heading, and page number, and where cross-cycle evolution is tracked through typed evolves_to edges. A hybrid retrieval layer combining Pinecone semantic search and Neo4j graph traversal answers plain-English questions with exact source citations in under 3 seconds. A two-stage LLM pipeline using step-back query decomposition followed by grounded synthesis with 7 specialist regulatory agent tools surfaces patterns that would otherwise require days of manual review.