Recently, our colleague Juan Sequeda published a technical breakdown of two methods aimed at making answering data questions easier: Data Co-pilots and Data Concierges. You can check out the full technical explanation here.  For a higher-level summary, read on. 

Navigating enterprise databases can feel like searching for a needle in a haystack, especially for business users who aren’t fluently using SQL. To help, let’s walk through two ways to make answering data questions easier – Data Co-Pilots and Data Concierges – and uncover which methodology might be best for you. 

Data co-pilot: Helping technical users navigate data

“Data co-pilot” is the marketing term for text-to-SQL. Modern data co-pilots act as translators between a technical user and the database. It converts plain-language questions into SQL queries, helping users quickly get the answers they need. When a user interacts with a co-pilot, it’s akin to asking a question to a person who has limited knowledge about the data and it returns code.

Co-pilots work best for tech-savvy users who are comfortable editing and understanding SQL queries, even if they're not always 100% accurate. The Data co-pilots of today are LLM-based.

The co-pilot approach isn't perfect, since accuracy drops with more complex database structures. However, this approach can be helpful for technical teams who can tweak queries for specific needs. While these tools can save time, they aren't the holy grail for broader business users.

Data concierge: Bringing answers to everyone

Business executives, marketing teams, and customers require more than just a co-pilot—they need a “Data Concierge,” which provides accurate answers that make sense to everyone. This tool functions like a knowledgeable advisor, leveraging a knowledge graph to understand the data and provide trustworthy insights.

The target users for data concierge are users in lines of business (executives, finance, sales, marketing) and consumers (the customers, lawyers, patients, etc). The users don’t have knowledge about what is going on underneath the hood, and they don’t need to.

A data concierge leverages a knowledge graph which is how it provides higher accuracy. Our benchmark research provided evidence that LLM accuracy when answering questions on enterprise SQL databases increased 3X with knowledge graphs. Investing in knowledge graphs and a data catalog, that provide the context of your organization, are foundational for a data concierge.

Unlike a co-pilot, which focuses solely on technical accuracy, a data concierge combines LLMs and traditional AI techniques to deliver answers that align with business goals. By the way, that’s what makes the AI Context Engine such a powerful tool for businesses. 

Which is right for you? 

While data co-pilots streamline data exploration for technical teams, data concierges democratize data access for a wider audience. They offer instant insights, enabling innovative decisions and eliminating the bottleneck of waiting on data teams.

Both solutions have their place in the data ecosystem. Co-pilots help technical users move quickly, while concierges provide transformative data access for everyone. By investing in these tools, businesses can make the once impossible, possible—for all teams across the organization.