Businesses turn to AI to query live operational data

Demand is rising for AI tools that let businesses query their own operational data in plain English, according to SyncHub, which linked the shift to growing use of AI assistants connected to live business systems.

The Auckland-based data integration provider said customers are moving away from relying solely on dashboards, static reports and exported spreadsheets. Instead, they are asking direct questions about sales, stock, staffing, invoices and performance through AI tools such as its own AI Assistant, as well as services including Claude and ChatGPT.

At the centre of that change is Model Context Protocol, or MCP, which allows AI tools to connect more directly with business systems and data sources. That gives users a way to retrieve answers from live operational data rather than from manually updated reports or disconnected file exports.

SyncHub argues this only works when data from multiple business systems has already been cleaned, structured and brought together. Many businesses use separate cloud platforms for accounting, customer management, human resources, inventory and timesheets, with each system storing information in different formats.

SyncHub positions itself as an aggregator across those systems. While individual software vendors are starting to offer their own MCP servers, those links are generally limited to each platform's own data rather than a combined view across the wider business.

That distinction has also shaped product development. In October, the group released a custom database connector designed to let customers query cloud software data alongside information held in on-premise systems, legacy databases and in-house software.

Ben Liebert, Founder, SyncHub, said the move followed customer demand from companies that needed access to data held outside standard SaaS products.

"Although businesses lean on SaaS platforms for the majority of their business software, many also have had to build custom software to meet their particular requirements. We had a lot of demand from customers that needed this data alongside their cloud data, so we built a connector for it," said Liebert.

He said the rise in custom database connections reflected wider changes in business software. AI-assisted coding tools have made it easier for companies to develop their own applications, creating more internal databases that still need to feed into reporting and analysis.

Shift in access

Liebert said customers increasingly want immediate answers rather than another reporting layer that still requires manual interpretation. He added that the quality of responses from AI systems still depends on the underlying state of the data they draw from.

"What we're seeing is that people don't want another layer of dashboards if they still have to interpret everything themselves. They want to ask a direct question about their business and get a reliable answer back. That only works if the data underneath is clean, structured and properly connected. AI isn't magic. If the data foundation is poor, the answers will be poor too," said Liebert.

He added that conventional reporting tools still have a place, especially for formal reporting and dashboard-based monitoring. But natural-language querying is becoming the quicker first step when someone simply needs an answer from business data.

"For years, the default process was to open a dashboard, export a file, or find someone technical to write a query," said Liebert. "MCP is speeding that up, because it lets businesses connect their structured data straight to the AI tools they're already using, whether that's our own AI Assistant or something like Claude."

Customer example

One customer cited by SyncHub is UK retailer Pooky. Benjamin Durrant, Business Solutions Architect, Pooky, said the company first used SyncHub to obtain structured data from its ERP system and then expanded into AI-driven querying.

"We joined SyncHub originally to achieve what seemed impossible: to get good structured data from our ERP," said Durrant. "Since then we have utilised the inbuilt AI to help create SQL queries, which we now use for our daily KPIs.

"More recently, we connected SyncHub to our Claude using the MCP connector, and it's made getting insights a thousand times easier. We now have reports for every area of our business, providing insights and tracking every day with little to no effort," said Durrant.

SyncHub said growth in use of its AI Assistant and MCP server suggests businesses are moving into a more practical phase of AI adoption. In that phase, the focus is less on generic prompting and more on linking AI systems to current business data so they can act as a direct interface to information already held across the organisation.

The company said it now serves more than 500 businesses across 30 countries.