# industrial-edge/ai-sdk-tutorials architecture

> Architecture diagram of the GitHub repository industrial-edge/ai-sdk-tutorials, generated by GitDiagram from its file tree, README and a sample of source files (last updated 2026-11-18). It is an AI-made overview: check details against the source.

- Interactive diagram: https://gitdiagram.com/industrial-edge/ai-sdk-tutorials
- Repository: https://github.com/industrial-edge/ai-sdk-tutorials

## Overview

This repository is a collection of AI SDK tutorials and guides, not the SDK implementation itself. Its principal user workflow is to prepare data, train a model, package it as an inference pipeline, test the package locally, and optionally generate an AI@Edge pipeline. The graph includes sampled inference examples and README-documented workflow capabilities; links to unsampled tutorial stages are documented rather than inferred from code.

Main entry points: [`e2e-tutorials`](https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/e2e-tutorials) (Pipeline Packaging), [`e2e-tutorials/state_identifier/src/si/inference.py`](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/state_identifier/src/si/inference.py) (State Inference), [`e2e-tutorials/image_classification/entrypoint.py`](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/entrypoint.py) (Image Entry Point), [`e2e-tutorials/image_classification/entrypoint_tf.py`](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/entrypoint_tf.py) (TensorFlow Entry)

## Diagram (Mermaid)

```mermaid
flowchart TD

subgraph group_workflow["ML Workflow"]
  node_training["Model Training"]
  node_packaging["Pipeline Packaging"]
  node_localtest["Local Testing"]
  node_edge["Edge Pipeline"]
end

subgraph group_examples["Inference Examples"]
  node_imageentry["Image Entry Point<br/>[entrypoint.py]"]
  node_tflite["TFLite Classifier"]
  node_tfentry["TensorFlow Entry<br/>[entrypoint_tf.py]"]
  node_tfclassifier["TensorFlow Classifier"]
  node_stateentry["State Entry Point<br/>[entrypoint.py]"]
  node_stateinfer["State Inference<br/>[inference.py]"]
  node_statepipeline["Time-Series Pipeline<br/>[pipeline.py]"]
  node_detpost["Detection Postprocess<br/>[postprocessing.py]"]
  node_detutils["Detection Drawing<br/>[image_utils.py]"]
  node_detpre["Detection Preprocess<br/>[preprocessing.py]"]
  node_segmentation["Image Segmentation<br/>[segmentation.py]"]
  node_batch["Batch State Example<br/>[pipeline.py]"]
  node_softsensor["Soft Sensor Example<br/>[entrypoint.py]"]
end

subgraph group_guides["Packaging Guides"]
  node_howtos["Packaging How-tos"]
  node_notebooks["Runnable How-to Notebooks"]
end

node_user(("Tutorial User"))

node_user -->|"prepares data"| node_training
node_training -->|"packages model"| node_packaging
node_packaging -->|"tests package"| node_localtest
node_packaging -->|"generates pipeline"| node_edge
node_user -->|"consults"| node_howtos
node_user -->|"runs examples"| node_notebooks
node_imageentry -->|"calls classifier"| node_tflite
node_tfentry -->|"calls classifier"| node_tfclassifier
node_stateentry -->|"delegates inference"| node_stateinfer
node_stateinfer -.->|"transforms window"| node_statepipeline
node_detpost -->|"draws predictions"| node_detutils

click node_training "https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/e2e-tutorials"
click node_packaging "https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/e2e-tutorials"
click node_localtest "https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/e2e-tutorials"
click node_edge "https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/e2e-tutorials"
click node_imageentry "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/entrypoint.py"
click node_tflite "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/src/vision_classifier.py"
click node_tfentry "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/entrypoint_tf.py"
click node_tfclassifier "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/src/vision_classifier_tf.py"
click node_stateentry "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/state_identifier/entrypoint.py"
click node_stateinfer "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/state_identifier/src/si/inference.py"
click node_statepipeline "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/state_identifier/src/si/pipeline.py"
click node_detpost "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/object_detection/src/postprocessing/postprocessing.py"
click node_detutils "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/object_detection/src/image_utils.py"
click node_detpre "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/object_detection/src/preprocessing/preprocessing.py"
click node_segmentation "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/segmentation/src/segmentation.py"
click node_batch "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/batch_state_identifier/src/si/pipeline.py"
click node_softsensor "https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/soft_sensor/src/inference/entrypoint.py"
click node_howtos "https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/howto-guides"
click node_notebooks "https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/howto-notebooks"

classDef toneNeutral fill:#f8fafc,stroke:#334155,stroke-width:1.5px,color:#0f172a
classDef toneBlue fill:#dbeafe,stroke:#2563eb,stroke-width:1.5px,color:#172554
classDef toneAmber fill:#fef3c7,stroke:#d97706,stroke-width:1.5px,color:#78350f
classDef toneMint fill:#dcfce7,stroke:#16a34a,stroke-width:1.5px,color:#14532d
classDef toneRose fill:#ffe4e6,stroke:#e11d48,stroke-width:1.5px,color:#881337
classDef toneIndigo fill:#e0e7ff,stroke:#4f46e5,stroke-width:1.5px,color:#312e81
classDef toneTeal fill:#ccfbf1,stroke:#0f766e,stroke-width:1.5px,color:#134e4a
class node_training,node_packaging,node_localtest,node_edge,node_user toneBlue
class node_imageentry,node_tflite,node_tfentry,node_tfclassifier,node_stateentry,node_stateinfer,node_statepipeline,node_detpost,node_detutils,node_detpre,node_segmentation,node_batch,node_softsensor toneAmber
class node_howtos,node_notebooks toneMint
```

## Components

### ML Workflow

- [Model Training](https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/e2e-tutorials): `e2e-tutorials`
- [Pipeline Packaging](https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/e2e-tutorials): `e2e-tutorials`
- [Local Testing](https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/e2e-tutorials): `e2e-tutorials`
- [Edge Pipeline](https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/e2e-tutorials): `e2e-tutorials`

### Inference Examples

- [Image Entry Point](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/entrypoint.py): `e2e-tutorials/image_classification/entrypoint.py`
- [TFLite Classifier](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/src/vision_classifier.py): `e2e-tutorials/image_classification/src/vision_classifier.py`
- [TensorFlow Entry](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/entrypoint_tf.py): `e2e-tutorials/image_classification/entrypoint_tf.py`
- [TensorFlow Classifier](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/image_classification/src/vision_classifier_tf.py): `e2e-tutorials/image_classification/src/vision_classifier_tf.py`
- [State Entry Point](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/state_identifier/entrypoint.py): `e2e-tutorials/state_identifier/entrypoint.py`
- [State Inference](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/state_identifier/src/si/inference.py): `e2e-tutorials/state_identifier/src/si/inference.py`
- [Time-Series Pipeline](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/state_identifier/src/si/pipeline.py): `e2e-tutorials/state_identifier/src/si/pipeline.py`
- [Detection Postprocess](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/object_detection/src/postprocessing/postprocessing.py): `e2e-tutorials/object_detection/src/postprocessing/postprocessing.py`
- [Detection Drawing](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/object_detection/src/image_utils.py): `e2e-tutorials/object_detection/src/image_utils.py`
- [Detection Preprocess](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/object_detection/src/preprocessing/preprocessing.py): `e2e-tutorials/object_detection/src/preprocessing/preprocessing.py`
- [Image Segmentation](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/segmentation/src/segmentation.py): `e2e-tutorials/segmentation/src/segmentation.py`
- [Batch State Example](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/batch_state_identifier/src/si/pipeline.py): `e2e-tutorials/batch_state_identifier/src/si/pipeline.py`
- [Soft Sensor Example](https://github.com/industrial-edge/ai-sdk-tutorials/blob/main/e2e-tutorials/soft_sensor/src/inference/entrypoint.py): `e2e-tutorials/soft_sensor/src/inference/entrypoint.py`

### Packaging Guides

- [Packaging How-tos](https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/howto-guides): `howto-guides`
- [Runnable How-to Notebooks](https://github.com/industrial-edge/ai-sdk-tutorials/tree/main/howto-notebooks): `howto-notebooks`

### Other components

- **Tutorial User**

## Connections

- Tutorial User → Model Training: prepares data (evidence: `README.md`)
- Model Training → Pipeline Packaging: packages model (evidence: `README.md`)
- Pipeline Packaging → Local Testing: tests package (evidence: `README.md`)
- Pipeline Packaging → Edge Pipeline: generates pipeline (evidence: `README.md`)
- Tutorial User → Packaging How-tos: consults (evidence: `README.md`)
- Tutorial User → Runnable How-to Notebooks: runs examples (evidence: `README.md`)
- Image Entry Point → TFLite Classifier: calls classifier (evidence: `e2e-tutorials/image_classification/entrypoint.py`)
- TensorFlow Entry → TensorFlow Classifier: calls classifier (evidence: `e2e-tutorials/image_classification/entrypoint_tf.py`)
- State Entry Point → State Inference: delegates inference (evidence: `e2e-tutorials/state_identifier/entrypoint.py`)
- State Inference → Time-Series Pipeline: transforms window
- Detection Postprocess → Detection Drawing: draws predictions (evidence: `e2e-tutorials/object_detection/src/postprocessing/postprocessing.py`)
