Why Your New AI Project is Actually a Data Project

Eliminating Data Silos, Connecting, and Organizing Data for AI

Managing data is becoming increasingly complex, and when you add AI to the mix, it adds another layer of complexity.

“Existing data often resides in multiple databases and data warehouses, which often contains duplicates, outliers, and irrelevant data points,” said Mona Chadha, director of category management at Amazon Web Services, in a recent Forbes article. “There are also gaps in the existing datasets. Organizations need better tools to clean and label the data. Poor data quality can result in inaccurate results and inconsistent model behavior, leading to lack of trust from customers and internal stakeholders,” said Chadha.

To give you some context, Salesforce® reports that the average enterprise has data in more than 800 applications, and only 29% of them are connected. “What’s more, it takes 35 applications on average to support just one customer interaction. It is no wonder that nine out of ten IT leaders say that these data silos create business challenges and impact customer experience,” explains the Salesforce AI Strategy Guide.

Recently, Salesforce introduced the next generation of its Einstein AI technology, which they unveiled at Dreamforce. Their AI Strategy Guide explains, “When integrated right into the flow of work, predictive and generative AI can help you make informed decisions based on business insights. It can make an immediate impact in almost every department.” For example, Einstein AI can help sales teams with tasks like researching accounts, preparing for meetings, and automatically updating account information in Salesforce.

But before organizations can experience the benefits of AI, they need to ensure their existing data assets are properly managed. In their study, “CDO Insights 2023: How to Empower Data-Led Business Resiliency,” Informatica reports, “Data continues to grow at an exponential rate — 55% of data leaders report more than 1,000 sources of data at their organizations.” Adding to this, the study notes that AI algorithms are also increasing in complexity.

“Every AI project should begin as a data project. And that data? It can’t live in silos. If it does, you risk lower quality outputs from your gen AI. The first important step is to connect, organize, and harmonize your data so you can understand and meet the needs of your customers with generative AI,” advises the AI Strategy Guide.

How Our Customers Are Managing and Connecting Their Salesforce Data

Located in Menlo Park, California, Swivl is a distributed team of technologists, educators, and innovators who are committed to improving the education system. As Swivl expanded, rolling up data daily became more time-consuming. They found that using Rollup Helper’s added targeted rollup functionality helps them manage their data more efficiently. Now they can easily collect product data at the Opportunity level, or even at the Account level.

The customer experience platform AskNicely uses Rollup Helper so they can get actionable, real-time data for reporting. “We probably have a dozen different separate rollups pulling in information from the daily stats that are identifying risks, use cases, and looking for specific product features,” said Sloan Morris, Director of Operations at AskNicely.

Preparing Your Data for AI: Solutions and Use Cases

In addition to overwhelming amounts of new data, old data in your Salesforce org can make it difficult for users to find the information they need.

One of the top Salesforce Data Storage consumers is outdated emails. You can overcome this by running a health check to identify outdated or unused data in your org. Once identified, data can be cleaned up using our Storage Helper app, which mass deletes data for standard or custom object(s). Storage Helper even backs up the Salesforce records that it deletes to a .csv file before the deletion occurs. This use case shows how you can Mass Delete Emails, Auto-Responses & Case Email Messages and save on Salesforce data storage costs.

Another way customers are improving their data and getting more insights is by using Lookup Helper to automatically relate Salesforce records. To keep data clean and organized, Lookup Helper can identify and eliminate duplicate records by looking for existing matches to the criteria. Lookup Helper can also auto-populate lookups using record matching rules, and group data by time period, geography, or any other category.

Rollup Helper elevates reporting with cross object rollups, and custom filters allow you to rollup data to a centralized object. For example, Rollup Helper gives you the ability to identify Names of Child Accounts on Parent Account or find information on the Number of Contacts without an Email Address.

Our new Data Quality Helper app ensures that your Salesforce users are entering or updating complete, correct, and non-duplicative data. This proactive approach is a great strategy for reducing or eliminating time spent on fixing and cleansing data later. To learn more, read Prevent Bad Data Creation and Find and Clean Up Existing Data.

When you’re struggling with decentralized data and disorganized reports, choosing the right Salesforce solution to manage all this information can be daunting. Sometimes a solution like Einstein will work best for your analytics needs, but other times a custom solution is needed.

Whether your data is in Salesforce or Excel, data in multiple locations needs to be refined, then accurate across the board and maintained. “We can create custom Visualforce pages — command centers for employees or analytics components — that allow you to see your information in a coherent manner,” said Mitchell Machor, Passage Technology’s Development Services owner.

Remember, before introducing your data to AI, the first step is to clean it, improve the quality, then make sure it stays in sync via automation or notification. Reach out to us to learn which analytics tools you should be building with and let us save you time and money.