Introduction

You have a pile of unstructured documents and need to answer questions that span them — "who works with people who worked on project X", "which vendors are connected to this incident". No single document contains the answer. RAG retrieval won't chain the facts for you. You need a knowledge graph: entities as nodes, typed relations as edges, so that multi-hop reasoning becomes graph traversal.

Building one used to mean training a named-entity recognizer on your domain, training a relation classifier, writing entity-resolution heuristics, and maintaining all three as your data shifted. With Claude, each of those stages becomes a prompt.

What You'll Learn

By the end of this guide you will be able to:

• Use structured outputs to extract typed entities and subject–predicate–object triples from arbitrary text with no training data
• Apply Claude-driven entity resolution to collapse surface-form variants into canonical nodes, replacing brittle string-similarity heuristics
• Assemble and query an in-memory graph, and run multi-hop questions by serializing subgraphs back to Claude
• Measure extraction quality with precision/recall against a gold set and reason about the cost/quality tradeoff between Haiku and Sonnet

Everything runs in memory with no database. The techniques transfer directly to Neo4j, Neptune, or a Postgres adjacency table when you need to scale.

Prerequisites

• Python 3.11+
• Anthropic API key (get one here)
• Basic familiarity with graphs (nodes, edges, traversal)

Setup

We use two models. Haiku handles the high-volume, schema-constrained extraction work where speed and cost matter more than nuance. Sonnet handles entity resolution and summarization, where the model needs to weigh conflicting evidence across documents.

import anthropic
client = anthropic.Anthropic()
Fast model for extraction
haiku = "claude-3-haiku-20240307"
Powerful model for resolution
sonnet = "claude-3-sonnet-20240229"

Building a Corpus

We need a handful of documents that talk about overlapping entities, so that entity resolution has real work to do. The Apollo program is a good test bed: six short Wikipedia summaries that all mention NASA, the Moon, several astronauts, and a launch vehicle — but each article names them slightly differently.

We fetch summaries from the Wikipedia REST API rather than full articles to keep token costs low. For a production pipeline you would chunk full documents; the extraction logic is identical.

Entity and Relation Extraction

Classical NER tags spans of text with labels (PERSON, ORG, LOC). Classical relation extraction then classifies pairs of spans into relation types. Both traditionally require labeled training data per domain.

We collapse both stages into a single Claude call per document. The key is structured outputs: we define the output shape as a Pydantic model and pass it to client.messages.parse(). Claude's response is guaranteed to validate against that schema and comes back as a typed Python object — no regex parsing, no JSON decode errors, no defensive isinstance checks.

from pydantic import BaseModel
from typing import List
class Entity(BaseModel):
    name: str
    type: str  # PERSON, ORG, LOC, EVENT, etc.
    description: str
class Relation(BaseModel):
    subject: str
    predicate: str
    object: str
class Extraction(BaseModel):
    entities: List[Entity]
    relations: List[Relation]
Extract from a document
def extract_entities_and_relations(text: str) -> Extraction:
    response = client.messages.parse(
        model=haiku,
        max_tokens=2000,
        system="You are a knowledge graph extraction system. Extract all named entities and their relationships from the text.",
        messages=[{"role": "user", "content": text}],
        response_model=Extraction
    )
    return response

Notice how the same real-world entity appears under different surface forms across documents — this is the entity resolution problem we solve next.

Entity Resolution

The raw extraction gives us overlapping mentions: "NASA" and "National Aeronautics and Space Administration", "Neil Armstrong" and "Armstrong", possibly "the Moon" and "Moon". If we build a graph directly from this, we get a fractured mess where the same concept is split across disconnected nodes.

Traditional approaches use string similarity (edit distance, Jaccard on tokens) plus blocking rules. That works for typos but fails on "Edwin Aldrin" vs "Buzz Aldrin" — two names with zero character overlap that refer to the same person.

We instead ask Claude to cluster entities of each type, using the one-line descriptions from extraction as disambiguation context. The descriptions matter: "Armstrong — first person to walk on the Moon" and "Armstrong — jazz trumpeter" have the same name but should not merge.

def resolve_entities(entities: List[Entity]) -> dict:
    """
    Returns a mapping from alias -> canonical name
    """
    # Group entities by type
    by_type = {}
    for e in entities:
        by_type.setdefault(e.type, []).append(e)
    
    alias_to_canonical = {}
    for etype, group in by_type.items():
        prompt = f"""
        Group these {etype} entities that refer to the same real-world entity.
        Use the descriptions for disambiguation.
        
        Entities:
        {chr(10).join(f'- {e.name}: {e.description}' for e in group)}
        
        Return a JSON mapping from each name to its canonical form.
        """
        response = client.messages.parse(
            model=sonnet,
            max_tokens=1000,
            system="You are an entity resolution system. Merge duplicate entities.",
            messages=[{"role": "user", "content": prompt}],
            response_model=dict
        )
        alias_to_canonical.update(response)
    
    return alias_to_canonical

Two failure modes to watch for. First, any raw name Claude leaves out of every cluster silently disappears from the graph — a production resolver should fall back to a single-element cluster for unmatched names so nothing is lost. Second, the resolver can over-merge: a specific mission like "Gemini 12" may get folded into the broader "Project Gemini" because the descriptions overlap. The first loses nodes, the second loses precision. Both are worth spot-checking in the output.

Assembling the Graph

With a clean alias map, we rewrite every relation endpoint to its canonical form and load the result into NetworkX. We use a MultiDiGraph because two entities can be connected by several distinct predicates ("launched from" and "operated by"), and direction matters ("Armstrong commanded Apollo 11" is not the same edge as "Apollo 11 commanded Armstrong").

Each node carries its type, the source document, and the original surface form for auditability.

import networkx as nx
def build_knowledge_graph(extractions: List[Extraction], alias_map: dict) -> nx.MultiDiGraph:
    G = nx.MultiDiGraph()
    
    for extraction in extractions:
        # Add nodes
        for entity in extraction.entities:
            canonical = alias_map.get(entity.name, entity.name)
            G.add_node(canonical, type=entity.type, description=entity.description)
        
        # Add edges
        for relation in extraction.relations:
            subj = alias_map.get(relation.subject, relation.subject)
            obj = alias_map.get(relation.object, relation.object)
            G.add_edge(subj, obj, predicate=relation.predicate)
    
    return G

Querying the Graph with Multi-Hop Reasoning

Now for the payoff: answering questions that require chaining facts across documents. We serialize the relevant subgraph back to Claude for reasoning.

def query_graph(G: nx.MultiDiGraph, question: str) -> str:
    # Serialize graph to text
    nodes_info = "\n".join(f"- {n} ({d['type']})" for n, d in G.nodes(data=True))
    edges_info = "\n".join(
        f"- {u} --[{d['predicate']}]--> {v}"
        for u, v, d in G.edges(data=True)
    )
    
    prompt = f"""
    Given this knowledge graph:
    
    Nodes:
    {nodes_info}
    
    Edges:
    {edges_info}
    
    Answer this question by traversing the graph: {question}
    """
    
    response = client.messages.create(
        model=sonnet,
        max_tokens=500,
        system="You are a knowledge graph query engine. Answer questions by reasoning over the graph.",
        messages=[{"role": "user", "content": prompt}]
    )
    return response.content[0].text

Measuring Quality

To trust your graph in production, you need to measure precision and recall against a gold standard. Create a small annotated dataset of documents with expected entities and relations, then compare your extraction against it.

def evaluate_extraction(gold: Extraction, predicted: Extraction):
    gold_entities = set((e.name, e.type) for e in gold.entities)
    pred_entities = set((e.name, e.type) for e in predicted.entities)
    
    true_positives = gold_entities & pred_entities
    precision = len(true_positives) / len(pred_entities) if pred_entities else 0
    recall = len(true_positives) / len(gold_entities) if gold_entities else 0
    
    return {"precision": precision, "recall": recall}

Cost/Quality Tradeoffs

Haiku is ~5x cheaper than Sonnet and faster, but may miss nuanced relations or struggle with ambiguous entity types. For high-volume extraction pipelines, use Haiku for initial passes and Sonnet for entity resolution and quality assurance on critical documents. A common pattern: extract with Haiku, resolve with Sonnet, and use Sonnet for any query that requires deep reasoning.

Key Takeaways

• No training data needed: Claude's structured outputs let you extract entities and relations with a single prompt, replacing traditional NER and relation classification pipelines.
• Claude beats string matching for entity resolution: Using descriptions for disambiguation handles cases like "Buzz Aldrin" vs "Edwin Aldrin" that edit distance cannot resolve.
• Multi-hop reasoning becomes graph traversal: Once your knowledge graph is built, complex questions that span documents become simple path queries.
• Monitor for failure modes: Watch for dropped entities (unmatched names) and over-merging (distinct concepts collapsed together). Always fall back to single-element clusters for unmatched names.
• Choose your model wisely: Use Haiku for high-volume extraction and Sonnet for resolution and reasoning to balance cost and quality.