QueryStory urges caution on AI-generated content

In 2009, Shapor Naghibzadeh discovered the power of compelling storytelling while working as a Google sysops engineer. During Operation Aurora, when state-sponsored hackers targeted the tech giant, he was summoned to a crisis room to clarify server activities, learning firsthand that verified knowledge is essential for tracing cyberattacks across fragmented networks—a process he found both expensive and slow. He believes large language models can replicate this verification capability across diverse databases in a fraction of the time.

For the next six years, Naghibzadeh focused on merging data and cybersecurity, leveraging Google’s infrastructure to create tools enabling security analysts to query complex datasets. In 2016, he co-founded Chronicle, a startup within Google’s X Labs, to extend these capabilities to external enterprises.

Last year, recognizing the growing role of large language models in data analysis, Naghibzadeh identified an opportunity to adapt his cybersecurity techniques for broader analytics applications. He co-founded QueryStory, launching from stealth today, alongside CTO Stanley Yang, a former Google colleague and lead engineer at EvolutionIQ, and CPO David Glusic, an Accenture veteran.

“Investigations follow a pattern: you ask the data a series of questions, then synthesize those answers into a coherent narrative,” Naghibzadeh explained. “That concept inspired the name QueryStory. It’s about constructing truthful, data-driven narratives.”

QueryStory secured a $6 million seed round in late 2025 from Brightmind Ventures and New York Life Ventures, valuing the company at $60 million. During this period, the startup has refined its product through customer pilots. Targeting large enterprises with extensive proprietary databases, QueryStory serves as a unified platform for data analysis and review, supporting teams such as sales and operations.

“We bridge the trust gap in AI by providing enterprises with actionable, reliable answers,” Naghibzadeh stated. “Rather than relying on human judgment or deploying armies of engineers, we have productized that expertise.”

Tim Del Bello, a partner at New York Life Ventures, invested in the company and is already using the platform to replace manual reporting processes, aiming to transform quarterly business reviews into real-time dashboards.

“This product is designed for decision-makers who need ground truth from complex, disparate data sources but lack dedicated data science or BI teams, particularly in highly regulated industries,” he told TechCrunch.

Solving the problem of brittle AI systems

I shared a database tracking space activity to help understand orbital operations by companies like SpaceX. QueryStory generated a visualization of this data within hours—a task that previously required a developer and several weeks. The platform produced sophisticated dashboards and, notably, included a confidence indicator explaining why the AI agents deemed the analyses accurate.

A visualization of satellites in orbit around the Earth.Image Credits:QueryStory / QueryStory

A set of queries derived from the space data.Image Credits:QueryStory / QueryStory

A QueryStory.Image Credits:QueryStory / QueryStory

QueryStory’s confidence indicator.Image Credits:QueryStory / QueryStory

While frontier labs offer co-working tools, these are intentionally restricted in user experience. QueryStory’s strategy assumes that enterprise users demand greater transparency, reliability, and control when integrating AI into their workflows.

For instance, a tech executive recently described using Claude Cowork to query a company database, then manually verifying the generated SQL queries before sending them to a data analyst for review. In QueryStory, these SQL queries appear automatically, and users can flag analyses for human review, with all feedback recorded directly on the platform.

“AI is more fragile than people realize when building durable systems that large-scale businesses depend on,” Tayler Sipperly, a partner at Brightmind Partners, told TechCrunch.

Naghibzadeh notes that connecting enterprise data to an LLM’s chat interface often results in hundreds or thousands of employees asking questions, generating conflicting versions of the truth, and creating unmanageable spreadsheets with no centralized source of record.

Economic factors also play a role. Although QueryStory is model-agnostic and currently utilizes the latest models from frontier labs, Naghibzadeh believes customers prefer service providers not incentivized to maximize intelligence consumption.

“We have a distinct advantage in not building our business around consuming compute, storage, or tokens,” Naghibzadeh said. He argues that purpose-built tools like QueryStory can achieve greater efficiency and accuracy than general-purpose agents by preserving context.

“What we sell is trust in the answers,” he said. “We deliver business value by empowering CFOs to understand exactly what these solutions will cost.”