Shapor Naghibzadeh learned the value of a good story in 2009 as a systems operations engineer at Google. When Chinese-backed hackers targeted the search giant as part of an effort dubbed “Operation Aurora,” he was called into a hastily assembled war room to explain what was happening on the company’s servers.
By tracking cyberattacks across disparate networks, Nagibzadeh learned the value of verified knowledge. However, it was an expensive and time-consuming task. He believes LLM can be used to quickly deploy the same functionality to any type of database.
Nagibzadeh spent the next six years focused on the relationship between data and cybersecurity, leveraging Google’s resources to build tools that enable security analysts to query complex data. In 2016, he co-founded a startup within Google’s X Labs called Chronicle that provides the same functionality to other companies.
Last year, as large-scale language models played a growing role in data analysis, Naghibzadeh saw new opportunities to apply the techniques he developed for cybersecurity to a variety of analyses. He co-founded and serves as CEO of QueryStory with CTO Stanley Yang, a former Google colleague and lead engineer at EvolutionIQ, and CPO David Glusic, an Accenture veteran. The startup emerged from stealth today.
“There’s this pattern in research: You ask a lot of questions about the data, and once you have a few questions, you put them together to create a story,” Naguibzadeh said. “That’s where the name QueryStory came from. It’s about using data to tell stories, right? It’s about putting together stories based on truth.”
QueryStory raised a $6 million seed round from Brightmind Ventures and New York Life Ventures at a $60 million valuation in late 2025, and has spent that time developing and piloting its product with customers. QueryStory is primarily targeted at large enterprises that manage large proprietary databases. It serves as a platform for integrating data analysis and reviews for users such as sales teams and operations managers.
“What we are doing is bridging the trust gap so that AI can provide actionable answers to businesses,” said Naghibzadeh. “Instead, we turned it into a product, borrowing human judgment and an army of forward-deployed engineers.”
Tim Del Bello, a partner at New York Life Ventures, invested in the company. He also uses the platform to replace the work of several people and create quarterly business reviews. Now we want it to be a real-time dashboard.
“This product was built for people like me, truth-seeking decision makers who need to work with complex and disparate data sources but don’t have a data science or BI team at their disposal, especially if they operate in highly regulated industries,” he told TechCrunch.
Solving problems with weak AI systems
I shared a database of space activities that helps companies like SpaceX understand what they’re doing in orbit. QueryStory created that data visualization in hours, whereas a project I once did with a developer took weeks. It produced sophisticated dashboards and analytics, and perhaps most notably, revealed confidence metrics that show why the AI agent believes the analysis is accurate.
This kind of work can be done using collaboration tools built by Frontier Labs, but these tools have an intentionally limited user experience. One thing QueryStory is betting on is that users, especially those in large enterprises, want more transparency, trust, and control when integrating AI into their workflows.
As an example, an executive at a tech company recently told TechCrunch about using Claude Cowork to query an internal database and asking the model to show him the SQL queries it created to make sure they made sense before sending them to data analysts for human review. QueryStory automatically displays these SQL queries and allows users to flag analyzes for review by colleagues, which are logged on the platform.
“If you want to build things that need to be durable and large businesses rely on AI, AI is more fragile than people think,” Tayler Sipperly, a partner at Brightmind Partners, told TechCrunch.
Naghibzadeh points out that when companies connect their data to LLM’s chat UI, “hundreds or thousands of people within the organization are asking questions, coming up with their own truths, and sharing them in slide decks. You end up with a ton of content and no real place to store the content associated with your data.”
We also have to deal with the economic aspects. QueryStory is built to be model agnostic, but currently primarily uses the latest models provided by Frontier Labs. Although his company competes with Frontier Labs on a product basis, Nagibzadeh believes customers will prefer working with service providers who are not incentivized to sell as much intelligence as possible.
“We’re accomplishing a lot here by not being a company that built its business around a consumption model of compute, storage, and tokens,” Nagibzadeh said. He argues that specialized tools like QueryStory can be more efficient and accurate than general-purpose agents by understanding and preserving context.
“What we’re selling is trust in the answer, right?” he said. “What we’re selling is the value we add to the business, and our ultimate goal is to help CFOs understand, ‘How much is this going to cost?'”
If you buy through links in our articles, we may earn a small commission. This does not affect editorial independence.
