Data Cloud
Pricing Use Case

Makana Health System connects several key data sources into Data Cloud to build a unified patient view:

• Salesforce Health Cloud: Patient relationship data including care coordination workflows, care team assignments, patient engagement history, outreach campaign participation, risk stratification scores, social determinants of health, care gaps, and service requests.
• Electronic Health Records (EHR): Clinical data including patient demographics, medical history, diagnoses, treatment plans, and medication lists from their primary EHR system.
• Clinical Treatment Guidelines: A centralized knowledge base of articles, stored in Salesforce knowledge, that provides comprehensive best practices and actionable instructions for patient care, encompassing medication protocols, post-treatment recovery recommendations, and crucial warning indicators.

To understand the credits and cost related to connecting and transforming this data we must first start by understanding the scope of profile and engagement data they’d like to bring in from these destinations which is outlined below:

Based on the data volumes and ingestion methods for Makana Health System:

• Profile Data: 10 million records. Based on this scope, we can determine the total amount of data to be processed in year one related to profile records:
  • From Salesforce Health Cloud— 1/3rd of profile records: 3.3 million profiles ingested with 5% updated daily, totaling roughly 64 million rows
  • From EHR system— 2/3rds of profile records: 6.7 million profiles ingested with 5% updated daily, totaling roughly 128 million rows
• Engagement Data: 30 million records annually inclusive of information about patient visits and communications, online portal activity, care plan adherence, and billing details.. Based on this scope, we can determine the total amount of data to be processed in year one related to engagement records, assuming Makana initially uploads one previous quarter’s worth of historical engagement data to start (represented by 25% of the total):
  • From Salesforce Health Cloud- 1/3rd of engagement records: 2.5 million historical records ingested initially with another 10 million records expected throughout the rest of year one, totaling 12.5 million rows annually
  • From EHR system- 2/3rds of engagement records: 5 million historical records ingested annually with another 20 million records expected through the rest of year one, totaling 25 million rows annually
• Data Transformations applied to 1% of all ingested data. In order to get the data into a proper format for harmonization, Makana runs data transforms on a small portion of the records ingested, every day processing a total of 173 million rows annually
• Unstructured Data Processed: 7,500 knowledge articles. Makana will process 3,821 MB of unstructured data annually for its Clinical Treatment Guidelines Repository. This estimate accounts for an average article size of 0.08 MB, plus a 2% monthly repository growth and 10% monthly article revisions.

Summary Table for Credits and Cost
The "Credits Consumed" column is calculated using the following formula: (Rows Processed * Credit Multiplier) / Unit of Multiplier

To create a single, actionable view of each customer, Makana unifies profiles from all connected data sources using Data Cloud's identity resolution capabilities. The number of rows processed for this action is calculated based on the number of source profiles processed by an identity resolution ruleset. After the first time a ruleset runs, only new or modified source profiles are counted. Here again, we've used the assumed rate of 5% of profiles updated daily, which results in an estimate of approximately 192 million profile-related records annually which results in 3.3 million unified profiles annually, with credit consumption based on the number of profiles unified.

Summary Table for Credits and Cost
The "Credits Consumed" column is calculated using the following formula: (Rows Processed * Credit Multiplier) / Unit of Multiplier

With unified patient profiles and relevant unstructured data fully indexed, Makana Health System activates its data using Agentforce to produce an always-on service agent that can assist patients with inquiries regarding post visit treatment recommendations. By grounding the agent in each patient's unified profile and leveraging a search retriever to access the Clinical Treatment Guidelines, it can deliver personalized recommendations and answers. This helps resolve patient cases faster and boosts satisfaction.

To achieve this, Agentforce will execute two specific data queries during patient conversations:

1. Data Query 1: Looking up the patient's unified profile. This action processes approximately 3.3 million records per query. With an annual estimate of 500 patient conversations daily (assuming roughly 5% of patients have one conversation per year), this totals 602 billion rows processed annually.
2. Data Query 2: Searching through clinical treatment guidelines. Scanning vectorized unstructured data for patient queries results in an estimated 42 billion rows processed annually. This annual projection is based on 500 conversations daily (5% of patients having a conversation that triggers this action), with each average query scanning 20% of the available unstructured data.

Combined, both data queries total 644 billion rows processed annually for this Agentforce activation.

Disclaimer: The pricing estimate does not include the costs for Agentforce Audit and Feedback logs (enabled by default in Data Cloud with Agentforce activation) or Einstein Requests.

Summary Table for Credits and Cost
The "Credits Consumed" column is calculated using the following formula: (Rows Processed * Credit Multiplier) / Unit of Multiplier

Based on the volume of unstructured data processed from various sources (clinical notes, medical images, EHR records, patient communications) to support the unified patient view and insights, Makana Health System will require roughly 2 TB of storage.

The estimated storage need of 2 TB for the first year is based on the substantial volume of laboratory data and provider notes, which are processed and stored in an optimized format within Data Cloud. This estimate also accounts for the storage of 3.3 million unified patient profiles with their associated metadata.

Summary Table for storage needs and costs