Practical notes: Introducing Docling Pipelines: Building Enterprise-Grade RAG

This walkthrough rebuilds the path from raw materials to a working system for: Introducing Docling Pipelines: Building Enterprise-Grade RAG Data Pipelines. The focus is operable steps, explicit checks, and code that you can drop into a repo without guessing intent.

The Problem With Glue Code

For the The Problem With Glue stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion. Cite the passages that actually grounded the answer. Without citations, operators cannot tell hallucination from an indexing gap.

What is docling-pipelines?

For the What is docling-pipelines stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments. Cite the passages that actually grounded the answer. Without citations, operators cannot tell hallucination from an indexing gap.

bash
pip install docling-pipelines

The Core Mental Model: Operators and Flows

For the The Core Mental Model stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Prefer structured outputs with schema validation over free-form prose when the next step is code or a tool call.

from docpipe.core.operators.abstract_operator import AbstractOperator, OperatorCategory
import pyarrow as pa

class MyOperator(AbstractOperator):
  short_name = "my_operator"
  category = OperatorCategory.Quality

  def __init__(self, config: dict) -> None:
    super().__init__(config)

  def transform(self, table: pa.Table) -> tuple[list[pa.Table], dict]:
    metadata = self.create_base_metadata(total_docs_count=len(table))
    # … process table …
    return [table], metadata

  @staticmethod
  def get_metadata() -> dict:
  return {"short_name": "my_operator", "description": "…"}

A Tour of the Operator Ecosystem

For the A Tour of the stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Cite the passages that actually grounded the answer. Without citations, operators cannot tell hallucination from an indexing gap.

What a Real Flow Looks Like

For the What a Real Flow stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Cite the passages that actually grounded the answer. Without citations, operators cannot tell hallucination from an indexing gap. For the What a Real Flow stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

{
  "flow_name": "complete-document-pipeline",
  "global_config": {
    "doc_column": "content",
    "storage": "in-memory"
  },
  "flow": [
    {
      "type": "ingest_source",
      "name": "ingest_local_folder",
      "config": {
        "provider": "filesystem",
        "connection_params": {
          "paths": [
            "./sample_documents"
          ]
        },
        "include_filter": "pdf,txt,docx"
      }
    },
    {
      "type": "extract_operator",
      "name": "extract_with_docling",
      "config": {
        "text_extraction": {
          "provider": "docling_library"
        },
        "entity_extraction": {
          "provider": "none"
        }
      },
      "depends_on": [
        "ingest_local_folder"
      ]
    },
    {
      "type": "chunker",
      "name": "simple_chunker",
      "config": {
        "chunk_type": "simple",
        "chunk_size": 512,
        "chunk_overlap": 50
      },
      "depends_on": [
        "extract_with_docling"
      ]
    },
    {
      "type": "embeddings",
      "name": "ollama_embeddings",
      "config": {
        "provider": "litellm",
        "provider_config": {
          "model_id": "openai/nomic-embed-text",
          "api_base": "http://localhost:11434/v1"
        },
        "embeddings_column": "embeddings"
      },
      "depends_on": [
        "simple_chunker"
      ]
    },
    {
      "type": "vectordb",
      "name": "opensearch_vector_store",
      "config": {
        "provider": "opensearch",
        "doc_id_column": "doc_id_hash",
        "embeddings_column": "embeddings",
        "provider_config": {
          "index_name": "sample-documents-index",
          "host": "localhost",
          "port": 9200
        }
      },
      "depends_on": [
        "ollama_embeddings"
      ]
    }
  ]
}

docling-pipelines - flow-file pipeline.json

Advanced Patterns: Branching, Quality Routing, and Multi-Stage Enrichment

When working through the Advanced Patterns Branching Quality stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Measure recall on a fixed question set before tuning prompts. Prompt churn rarely fixes a weak retrieval surface.

{
  "name": "quality_branching",
  "type": "branching",
  "config": {
    "branches": {
      "high_quality_branch": {
        "link_name": "High Quality Documents",
        "criteria_json": {
          "criteria_list": [
            {
              "variable": "flesch_reading_ease",
              "operator": ">",
              "value": 60
            }
          ],
          "logical_operator": "AND"
        }
      },
      "low_quality_branch": {
        "link_name": "Low Quality Documents",
        "criteria_json": {
          "criteria_list": [
            {
              "variable": "flesch_reading_ease",
              "operator": "<=",
              "value": 60
            }
          ],
          "logical_operator": "AND"
        }
      }
    }
  },
  "depends_on": [
    "readability"
  ]
}

lang_detect → ededup → doc_quality → readability → sql_filter → embeddings

{
  "criteria_list": [
    {
      "variable": "docq_total_words",
      "operator": ">",
      "value": 100
    },
    {
      "variable": "lang_name",
      "operator": "=",
      "value": "en"
    }
  ],
  "logical_operator": "AND"
}

Three Ways to Run It

When working through the Three Ways to Run stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Measure recall on a fixed question set before tuning prompts. Prompt churn rarely fixes a weak retrieval surface.

docling-pipelines - flow-file pipeline.json
docling-pipelines - flow-file pipeline.json - validate # validate without running
docling-pipelines - list-operators - verbose # discover all registered operators

from docpipe.lib.docpipe_flow_manager import DocpipeFlowManager
manager = DocpipeFlowManager(flow_file="pipeline.json")
manager.execute()

uvicorn docpipe.api.main:app - host 0.0.0.0 - port 8000

Extending with Custom Operators

When working through the Extending with Custom Operators stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Measure recall on a fixed question set before tuning prompts. Prompt churn rarely fixes a weak retrieval surface. When working through the Extending with Custom Operators stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

DOCPIPE_CUSTOM_OPERATORS=./my_operators docling-pipelines - flow-file pipeline.json

Operational Excellence Built In

The Operational Excellence Built In stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph. Separate chunking policy from retrieval policy. Changing one should not force a rewrite of the other when quality metrics move.

The Docling Pipelines Series

The The Docling Pipelines Series stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish. Separate chunking policy from retrieval policy. Changing one should not force a rewrite of the other when quality metrics move.

What’s Next

The What s Next stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline. Separate chunking policy from retrieval policy. Changing one should not force a rewrite of the other when quality metrics move. The What s Next stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

Operational checklist

The Operational checklist stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope.

Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion.

Separate chunking policy from retrieval policy. Changing one should not force a rewrite of the other when quality metrics move.

Add a smoke test that exercises the critical path in CI with fixtures, not live paid APIs, whenever budgets allow.

Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

Separate chunking policy from retrieval policy. Changing one should not force a rewrite of the other when quality metrics move.

Before promoting the stack, freeze versions, capture a golden transcript for the critical path, and confirm rollback steps. Shared environments need rate limits, tenancy checks, and a clear owner for secret rotation. Prefer boring reliability over clever one-off demos.

Batch note for d05b0d271243: keep provider keys out of the repo, set a per-session token ceiling, and store transcripts next to the eval fixtures so later model swaps stay comparable.

When working through the hardening note 0 stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion.

Hardening detail 0/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

The hardening note 1 stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph.

Hardening detail 1/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

For the hardening note 2 stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

Hardening detail 2/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

When working through the hardening note 3 stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

Hardening detail 3/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

The hardening note 4 stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish.

Hardening detail 4/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

For the hardening note 5 stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion.

Hardening detail 5/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

When working through the hardening note 6 stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph.

Hardening detail 6/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

The hardening note 7 stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

Hardening detail 7/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

For the hardening note 8 stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

Hardening detail 8/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

When working through the hardening note 9 stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish.

Hardening detail 9/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

The hardening note 10 stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Treat this stage as a contract between inputs and validated outputs. Name the artifacts, define success checks, and refuse silent partial completion.

Hardening detail 10/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

For the hardening note 11 stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Keep configuration outside application code. Environment files, secret stores, and feature flags belong in one place operators can audit without reading the whole graph.

Hardening detail 11/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

When working through the hardening note 12 stage, write down the contract first: required inputs, success signal, and what happens on partial failure. That checklist keeps later code changes honest. Prefer small, testable units over sprawling scripts. When a step fails, the failure should point at a single responsibility rather than a tangled pipeline.

Hardening detail 12/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

The hardening note 13 stage works best when treated as a measurable surface. Capture one golden transcript, one failure case, and the rollback note before expanding scope. Record timings and token or query cost next to functional results. Cost visibility early prevents surprise bills when the path moves from demo to shared environments.

Hardening detail 13/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.

For the hardening note 14 stage, define the inputs, the owner of the step, and the exit criteria before changing code. Operators should be able to re-run the step from a known checkpoint without guessing hidden state. Document the happy path and the recovery path together. Retries, human gates, and dead-letter handling are part of the product, not later polish.

Hardening detail 14/754: measure wall time, error class, and token spend for this note, then decide whether to keep the change based on a fixed question set rather than anecdote.