SHAPINg the future with
HYPERAUTOMATION
Process intelligence + Automation
SHAPINg the future with
HYPERAUTOMATION
Process intelligence + Automation
IMPROVE PROCESS EXCELLENCE AND ACHIEVE SUSTAINABLE ADVANTAGES
Embarking on a hyperautomation journey involves a strategic, phased approach that ensures a smooth transition and maximization of benefits.
We understand that embarking on a hyperautomation journey is a pivotal step for organizations looking to enhance their operational efficiency and drive innovation. To successfully reach the summit of hyperautomation, our 3-step model is the perfect travel guide.
SHAPINg the future with
HYPERAUTOMATION
Process intelligence + Automation
IMPROVE PROCESS EXCELLENCE AND ACHIEVE SUSTAINABLE ADVANTAGES
Embarking on a hyperautomation journey involves a strategic, phased approach that ensures a smooth transition and maximization of benefits.
We understand that embarking on a hyperautomation journey is a pivotal step for organizations looking to enhance their operational efficiency and drive innovation. To successfully reach the summit of hyperautomation, our 3-step model is the perfect travel guide.
STEP 1: DISCOVERY + ANALYSIS
.. provides the foundation for a successful automation. It is crucial for identifying processes and use cases suitable for automation with focus on business value and a secure and scalable architecture.
Process Mining can optionally be used for a holistic view on your processes.
STEP 1: DISCOVERY + ANALYSIS
.. provides the foundation for a successful automation. It is crucial for identifying processes and use cases suitable for automation with focus on business value and a secure and scalable architecture.
Process Mining can optionally be used for a holistic view on your processes.
STEP 2: AUTOMATION
.. is all about applying the right tools and architecture in an incremental design-build-deploy cycle to create a tailored automation solution that fit your unique requirements with measurable business impact.
RPA (Robotic Process Automation): Ideal for automating repetitive, rule-based tasks such as data entry, system updates, and file transfers. RPA is best used where structured data and stable processes are present, helping reduce manual errors and operational costs.
Agents (Intelligent Agentic Automation): AI agents, also kown as intelligent agents, are software or hardware entities that autonomously perform tasks, process data, make trustworthy decisions and interact with their environment or other agents. These agents are based on advanced algorithms and machine learning that enable them to adapt to changing conditions and constantly learn.
IDP (Intelligent Document Processing): Enables automation of document-centric tasks by extracting, classifying, and validating unstructured data from sources like PDFs, invoices, and forms. IDP can be applied in all industries and processes where large volumes of documents are processed.
API (Reusable Components): Facilitates seamless integration across systems, applications, and services. APIs are critical when e.g. multiple platforms or agents need to exchange data securly and in real-time, enabling end-to-end automation and enhanced digital experiences.
Each of these tools can be deployed individually or in combination, depending on the complexity and nature of your processes. The result is a streamlined, intelligent automation solution that enhances productivity, reduces costs, and improves overall operational agility.
STEP 2: AUTOMATION
.. is all about applying the right tools and architecture in an incremental design-build-deploy cycle to create a tailored automation solution that fit your unique requirements with measurable business impact.
RPA (Robotic Process Automation): Ideal for automating repetitive, rule-based tasks such as data entry, system updates, and file transfers. RPA is best used where structured data and stable processes are present, helping reduce manual errors and operational costs.
Agents (Intelligent Agentic Automation): AI agents, also kown as intelligent agents, are software or hardware entities that autonomously perform tasks, process data, make trustworthy decisions and interact with their environment or other agents. These agents are based on advanced algorithms and machine learning that enable them to adapt to changing conditions and constantly learn.
IDP (Intelligent Document Processing): Enables automation of document-centric tasks by extracting, classifying, and validating unstructured data from sources like PDFs, invoices, and forms. IDP can be applied in all industries and processes where large volumes of documents are processed.
API (Reusable Components): Facilitates seamless integration across systems, applications, and services. APIs are critical when e.g. multiple platforms or agents need to exchange data securly and in real-time, enabling end-to-end automation and enhanced digital experiences.
Each of these tools can be deployed individually or in combination, depending on the complexity and nature of your processes. The result is a streamlined, intelligent automation solution that enhances productivity, reduces costs, and improves overall operational agility.
STEP 3: SCALE + SUSTAIN
To roll out intelligent automation sustainably and securely in the company, it is important to scale effectively via value-creating use cases and to establish an architecture and governance that takes into account the “pay as you go” license models on the one hand and keeps pace with the rapidly changing environment on the other.
STEP 3: SCALE + SUSTAIN
To roll out intelligent automation sustainably and securely in the company, it is important to scale effectively via value-creating use cases and to establish an architecture and governance that takes into account the “pay as you go” license models on the one hand and keeps pace with the rapidly changing environment on the other.
SUCCESS STORIES
Summary
Step 1
Step 2
Step 3
Technologies & Tools Used
Key Functional Highlights
Summary
This success story describes an automated workflow designed for a service person who receives purchase orders (POs) from B2B customers in scanned PDF format. These POs are placed in a shared FTP folder. The service agent requests AgentForce to upload and process these documents. The automation includes uploading, document classification, data extraction, validation, and final order creation in Salesforce.
Step 1
The process begins when a service agent initiates an order upload request using AgentForce, a conversational interface. Upon receiving the request, AgentForce triggers an automated Salesforce Flow to handle the order intake process.
Step 1: Uploading PDFs from FTP to Salesforce
The first action within the flow is to call MuleSoft API #1, also known as the Uploader. This API connects to the shared FTP folder, where it reads the scanned PDF documents submitted by B2B customers. Since these files are typically not machine-readable, the API first converts them into Base64 format — a requirement for Salesforce document ingestion. It then uses the Salesforce REST API to upload each document into the system, returning a list of Content Version IDs that uniquely identify each uploaded file in Salesforce.
Step 1: Uploading PDFs from FTP to Salesforce
The first action within the flow is to call MuleSoft API #1, also known as the Uploader. This API connects to the shared FTP folder, where it reads the scanned PDF documents submitted by B2B customers. Since these files are typically not machine-readable, the API first converts them into Base64 format — a requirement for Salesforce document ingestion. It then uses the Salesforce REST API to upload each document into the system, returning a list of Content Version IDs that uniquely identify each uploaded file in Salesforce.
Step 2
Step 2: Document Classification and Data Extraction
The Salesforce Flow then loops through each uploaded document using the returned IDs. For every document, it constructs a download link and sends this to MuleSoft API #2, referred to as the IDP Processor. This second API retrieves the document from Salesforce and sends it to an Intelligent Document Processing (IDP) system.
The IDP performs two critical tasks:
1. Classification: It determines the document’s format and origin, which is important because each customer may use a different layout.
2. Data Extraction: It extracts key information from the scanned image, such as the purchase order type, order date, billing and shipping addresses, line items (including descriptions, unit prices, and quantities), the order total, buyer name, and other relevant fields. The extracted data is returned as a structured JSON object to the Salesforce Flow.
The Salesforce Flow then loops through each uploaded document using the returned IDs. For every document, it constructs a download link and sends this to MuleSoft API #2, referred to as the IDP Processor. This second API retrieves the document from Salesforce and sends it to an Intelligent Document Processing (IDP) system.
The IDP performs two critical tasks:
1. Classification: It determines the document’s format and origin, which is important because each customer may use a different layout.
2. Data Extraction: It extracts key information from the scanned image, such as the purchase order type, order date, billing and shipping addresses, line items (including descriptions, unit prices, and quantities), the order total, buyer name, and other relevant fields. The extracted data is returned as a structured JSON object to the Salesforce Flow.
Step 3
Step 3: Creating the Order in Salesforce
Once the data is received, the Salesforce Flow parses the JSON to extract both header and line-item details. It first creates the Order Header in Salesforce, then iterates through each product line to create corresponding Order Line Items. The original scanned PDF is also linked to the order record for reference. Once the creation process is complete, the flow returns the Order ID and Order Number to AgentForce.
Optional Post-Processing
After the order is created, the agent can use AgentForce to perform follow-up actions, such as checking the order status, reviewing billing or shipping details, or viewing related documents — all without leaving the conversational interface.
Once the data is received, the Salesforce Flow parses the JSON to extract both header and line-item details. It first creates the Order Header in Salesforce, then iterates through each product line to create corresponding Order Line Items. The original scanned PDF is also linked to the order record for reference. Once the creation process is complete, the flow returns the Order ID and Order Number to AgentForce.
Optional Post-Processing
After the order is created, the agent can use AgentForce to perform follow-up actions, such as checking the order status, reviewing billing or shipping details, or viewing related documents — all without leaving the conversational interface.
Technologies & Tools Used
| Component | Technology/Tool | Purpose |
|---|---|---|
| Order Storage | Salesforce | CRM system to store and manage orders |
| Automation Flow | Salesforce Flow (Auto-launched) | Orchestrates the end-to-end process |
| API Integration Layer | MuleSoft | Interfaces with FTP, IDP, and Salesforce APIs |
| Document Access | FTP Server | Source of scanned PDF purchase orders |
| Document Upload API | MuleSoft API #1 (Uploader) | Uploads PDFs to Salesforce |
| Content Management | Salesforce ContentVersion API | Stores and identifies documents |
| Intelligent Processing | IDP (Intelligent Document Processing) | AI-powered OCR & document understanding |
| Data Format Transformation | Base64 encoding | Required for document upload to Salesforce |
| Agent Interface | AgentForce | Interface used by service agents to trigger automation |
| Data Exchange | JSON | Format for structured data returned from IDP to Salesforce |
Key Functional Highlights
-
Scanned Image to Structured Data: Handles completely non-machine-readable PDFs using AI-powered document understanding.
Customer-Specific Variations: Automatically classifies document types before extracting data, accommodating varying customer formats.
No Hardcoded Prompts: Uses intelligent parsing to auto-identify fields like billing/shipping addresses, even in low-quality images.
Full Automation Loop: From scanned file to structured order with zero manual data entry.
This use case demonstrates a powerful combination of Salesforce, MuleSoft, and IDP AI to fully automate the intake of scanned B2B orders into a structured CRM workflow. It significantly reduces manual effort, improves accuracy, and provides end-to-end traceability from the original document to the final order record.
SUCCESS STORIES
Summary
Step 1
Step 2
Step 3
Technologies & Tools Used
Key Functional Highlights
Summary
This success story describes an automated workflow designed for a service person who receives purchase orders (POs) from B2B customers in scanned PDF format. These POs are placed in a shared FTP folder. The service agent requests AgentForce to upload and process these documents. The automation includes uploading, document classification, data extraction, validation, and final order creation in Salesforce.
Step 1
The process begins when a service agent initiates an order upload request using AgentForce, a conversational interface. Upon receiving the request, AgentForce triggers an automated Salesforce Flow to handle the order intake process.
Step 1: Uploading PDFs from FTP to Salesforce
The first action within the flow is to call MuleSoft API #1, also known as the Uploader. This API connects to the shared FTP folder, where it reads the scanned PDF documents submitted by B2B customers. Since these files are typically not machine-readable, the API first converts them into Base64 format — a requirement for Salesforce document ingestion. It then uses the Salesforce REST API to upload each document into the system, returning a list of Content Version IDs that uniquely identify each uploaded file in Salesforce.
Step 1: Uploading PDFs from FTP to Salesforce
The first action within the flow is to call MuleSoft API #1, also known as the Uploader. This API connects to the shared FTP folder, where it reads the scanned PDF documents submitted by B2B customers. Since these files are typically not machine-readable, the API first converts them into Base64 format — a requirement for Salesforce document ingestion. It then uses the Salesforce REST API to upload each document into the system, returning a list of Content Version IDs that uniquely identify each uploaded file in Salesforce.
Step 2
Step 2: Document Classification and Data Extraction
The Salesforce Flow then loops through each uploaded document using the returned IDs. For every document, it constructs a download link and sends this to MuleSoft API #2, referred to as the IDP Processor. This second API retrieves the document from Salesforce and sends it to an Intelligent Document Processing (IDP) system.
The IDP performs two critical tasks:
1. Classification: It determines the document’s format and origin, which is important because each customer may use a different layout.
2. Data Extraction: It extracts key information from the scanned image, such as the purchase order type, order date, billing and shipping addresses, line items (including descriptions, unit prices, and quantities), the order total, buyer name, and other relevant fields. The extracted data is returned as a structured JSON object to the Salesforce Flow.
The Salesforce Flow then loops through each uploaded document using the returned IDs. For every document, it constructs a download link and sends this to MuleSoft API #2, referred to as the IDP Processor. This second API retrieves the document from Salesforce and sends it to an Intelligent Document Processing (IDP) system.
The IDP performs two critical tasks:
1. Classification: It determines the document’s format and origin, which is important because each customer may use a different layout.
2. Data Extraction: It extracts key information from the scanned image, such as the purchase order type, order date, billing and shipping addresses, line items (including descriptions, unit prices, and quantities), the order total, buyer name, and other relevant fields. The extracted data is returned as a structured JSON object to the Salesforce Flow.
Step 3
Step 3: Creating the Order in Salesforce
Once the data is received, the Salesforce Flow parses the JSON to extract both header and line-item details. It first creates the Order Header in Salesforce, then iterates through each product line to create corresponding Order Line Items. The original scanned PDF is also linked to the order record for reference. Once the creation process is complete, the flow returns the Order ID and Order Number to AgentForce.
Optional Post-Processing
After the order is created, the agent can use AgentForce to perform follow-up actions, such as checking the order status, reviewing billing or shipping details, or viewing related documents — all without leaving the conversational interface.
Once the data is received, the Salesforce Flow parses the JSON to extract both header and line-item details. It first creates the Order Header in Salesforce, then iterates through each product line to create corresponding Order Line Items. The original scanned PDF is also linked to the order record for reference. Once the creation process is complete, the flow returns the Order ID and Order Number to AgentForce.
Optional Post-Processing
After the order is created, the agent can use AgentForce to perform follow-up actions, such as checking the order status, reviewing billing or shipping details, or viewing related documents — all without leaving the conversational interface.
Technologies & Tools Used
| Component | Technology/Tool | Purpose |
|---|---|---|
| Order Storage | Salesforce | CRM system to store and manage orders |
| Automation Flow | Salesforce Flow (Auto-launched) | Orchestrates the end-to-end process |
| API Integration Layer | MuleSoft | Interfaces with FTP, IDP, and Salesforce APIs |
| Document Access | FTP Server | Source of scanned PDF purchase orders |
| Document Upload API | MuleSoft API #1 (Uploader) | Uploads PDFs to Salesforce |
| Content Management | Salesforce ContentVersion API | Stores and identifies documents |
| Intelligent Processing | IDP (Intelligent Document Processing) | AI-powered OCR & document understanding |
| Data Format Transformation | Base64 encoding | Required for document upload to Salesforce |
| Agent Interface | AgentForce | Interface used by service agents to trigger automation |
| Data Exchange | JSON | Format for structured data returned from IDP to Salesforce |
Key Functional Highlights
-
Scanned Image to Structured Data: Handles completely non-machine-readable PDFs using AI-powered document understanding.
Customer-Specific Variations: Automatically classifies document types before extracting data, accommodating varying customer formats.
No Hardcoded Prompts: Uses intelligent parsing to auto-identify fields like billing/shipping addresses, even in low-quality images.
Full Automation Loop: From scanned file to structured order with zero manual data entry.
This use case demonstrates a powerful combination of Salesforce, MuleSoft, and IDP AI to fully automate the intake of scanned B2B orders into a structured CRM workflow. It significantly reduces manual effort, improves accuracy, and provides end-to-end traceability from the original document to the final order record.
"For us, transforming your business is not just a project, but a joint journey based on reliable partnership!"
Tobias Stein
Director
Process Intelligence + automation
tobias.stein [at] lpdg.io
"To master the leap from heavyweight transactional processes to intelligent and seamless business processes, we need to start with the most important part of the puzzle - people."
"For us, transforming your business is not just a project, but a joint journey based on reliable partnership!"
Tobias Stein
Director
Process Intelligence + automation
tobias.stein [at] lpdg.io
