The LLM Enrichment stage uses language models to extract structured data from document content. It can identify entities, classify content, extract key information, and generate structured outputs.
Stage Category : APPLY (Enriches documents)Transformation : N documents → N documents (with extracted data added)
When to Use Use Case Description Entity extraction Extract names, dates, amounts from text Content classification Categorize documents by topic/type Key information extraction Pull specific facts from unstructured text Structured output generation Convert prose to structured data
When NOT to Use Scenario Recommended Alternative Simple field transformation json_transformPredefined taxonomy classification taxonomy_enrichLarge-scale processing Pre-process during indexing Real-time low-latency Use cached extractions
Parameters Parameter Type Default Description modelstring Required LLM model to use promptstring Required Extraction instructions content_fieldstring contentField to analyze output_fieldstring extractedField for extracted data output_schemaobject nullJSON schema for structured output batch_sizeinteger 5Documents per LLM call
Available Models Model Speed Quality Best For gpt-4o-miniFast Good Simple extractions gpt-4oMedium Excellent Complex analysis claude-3-haikuFast Good Quick processing claude-3-sonnetMedium Excellent Nuanced extraction
Configuration Examples Basic Entity Extraction
Structured Output with Schema
Key Facts Extraction
Topic Classification
Contact Information Extraction
{ "stage_type" : "apply" , "stage_id" : "llm_enrichment" , "parameters" : { "model" : "gpt-4o-mini" , "prompt" : "Extract all company names and person names mentioned in this document." , "output_field" : "entities" } }
Output Schema Define structured output using JSON Schema:
{ "output_schema" : { "type" : "object" , "properties" : { "field_name" : { "type" : "string" }, "numeric_field" : { "type" : "number" }, "boolean_field" : { "type" : "boolean" }, "array_field" : { "type" : "array" , "items" : { "type" : "string" }}, "enum_field" : { "type" : "string" , "enum" : [ "option1" , "option2" , "option3" ]} }, "required" : [ "field_name" ] } }
Supported Types Type Description stringText values numberNumeric values booleanTrue/false arrayLists of items objectNested objects
Output Examples Without Schema { "document_id" : "doc_123" , "content" : "Apple Inc. announced..." , "entities" : "Companies: Apple Inc., Microsoft \n People: Tim Cook, Satya Nadella" }
With Schema { "document_id" : "doc_123" , "content" : "Great product, 5 stars!..." , "review_analysis" : { "sentiment" : "positive" , "rating_mentioned" : 5 , "pros" : [ "easy to use" , "great value" , "fast shipping" ], "cons" : [ "packaging could be better" ], "would_recommend" : true } }
Metric Value Latency 300-800ms per batch Batch size 5 documents default Token usage ~200 tokens per document Parallel Batches processed concurrently
LLM enrichment is expensive. Consider pre-computing extractions during indexing for frequently accessed data.
Common Pipeline Patterns [ { "stage_type" : "filter" , "stage_id" : "semantic_search" , "parameters" : { "query" : "{{INPUT.query}}" , "vector_index" : "text_extractor_v1_embedding" , "top_k" : 20 } }, { "stage_type" : "apply" , "stage_id" : "llm_enrichment" , "parameters" : { "model" : "gpt-4o-mini" , "prompt" : "Extract the main topic and sentiment." , "output_field" : "analysis" , "output_schema" : { "type" : "object" , "properties" : { "topic" : { "type" : "string" }, "sentiment" : { "type" : "string" , "enum" : [ "positive" , "neutral" , "negative" ]} } } } }, { "stage_type" : "filter" , "stage_id" : "structured_filter" , "parameters" : { "conditions" : { "field" : "analysis.sentiment" , "operator" : "eq" , "value" : "positive" } } } ]
[ { "stage_type" : "filter" , "stage_id" : "hybrid_search" , "parameters" : { "query" : "{{INPUT.query}}" , "vector_index" : "text_extractor_v1_embedding" , "top_k" : 10 } }, { "stage_type" : "apply" , "stage_id" : "llm_enrichment" , "parameters" : { "model" : "gpt-4o" , "prompt" : "Extract all entities with their types and relationships." , "output_field" : "entities" , "output_schema" : { "type" : "object" , "properties" : { "people" : { "type" : "array" , "items" : { "type" : "object" , "properties" : { "name" : { "type" : "string" }, "role" : { "type" : "string" }}}}, "organizations" : { "type" : "array" , "items" : { "type" : "string" }}, "locations" : { "type" : "array" , "items" : { "type" : "string" }}, "dates" : { "type" : "array" , "items" : { "type" : "string" }} } } } } ]
Writing Effective Prompts Good Prompts ✓ "Extract the product name, price, and key features from this product listing." ✓ "Identify all dates mentioned and their associated events." ✓ "Classify the sentiment as positive, neutral, or negative, and explain why."
Poor Prompts ✗ "Analyze this document" (too vague) ✗ "Get the data" (not specific) ✗ "Tell me about it" (unclear output)
Be specific about what to extract and in what format. When using output_schema, the LLM will conform to the structure.
Error Handling Error Behavior LLM timeout Retry once, then null result Schema validation fail Raw text in output_field Rate limit Automatic backoff Empty content Skip enrichment
