Shapor Naghibzadeh’s journey at Google during Operation Aurora taught him the high cost of chasing elusive truths in complex data systems. He’s now applying those lessons through QueryStory, a startup designed to give enterprises confidence in the answers generated by large language models (LLMs). QueryStory officially emerged from stealth today.
From cybersecurity to narrative-driven AI
Naghibzadeh’s career has spanned work in Google’s systems operations, where he traced attacks like the 2009 China-backed Operation Aurora, and later co-founding Chronicle to help companies make sense of security data. In late 2025, he and co-founders Stanley Yang (former Google and EvolutionIQ engineer) and David Glusic (ex-Accenture) launched QueryStory. Based in Silicon Valley, QueryStory raised $6 million in seed funding from Brightmind Ventures and New York Life Ventures, valuing the startup at $60 million by the round’s close. Since then, it’s been developing its product and running pilot programs with enterprise users.
Bridging the trust gap with transparency
QueryStory positions itself as a tool for large organizations that juggle sprawling, proprietary data systems and need reliable insights without the overhead of full data science or business intelligence teams. Sales leaders, operations managers, even executives in regulated industries are the target users. What sets the platform apart is a blend of narrative generation, transparent model outputs (such as surfacing SQL queries automatically), and a visual “confidence indicator” that explains why a given insight is believed to be accurate. Users can also tag analyses for human review, and all these edits are tracked. This transparency aims to tackle the brittleness of AI systems, especially in enterprise settings.
Where other AI agents are general purpose, often opaque, and built around high compute or data costs, QueryStory aims to be model-agnostic. The startup says it doesn’t push usage-heavy compute or storage agendas. Rather, its goals are built around enabling decision-makers to understand both what an insight means and what it costs—or whether it can be trusted. CEO Naghibzadeh often frames the value offering as “trust in the answers,” especially in terms of cost visibility for roles like CFOs.
Proven value, but still early days
One example: when given a space-activity dataset—typically requiring weeks of work using developers—QueryStory delivered dashboards, visualizations, and analysis in a few hours, along with its confidence estimates. Also, investors like Tim Del Bello believe QueryStory is useful for decision-makers who need raw, trustworthy data narratives without BI teams. That said, the platform is still effectively in early deployment and validation with pilots rather than large-scale, entrenched use.
Economically, QueryStory competes with AI frontier labs and BI tools. Unlike some rivals whose business models depend heavily on token use, compute units, or storage costs, QueryStory claims it’s operating under a consumption-light model. The premise: customers will prefer paying for reliable, accurate insights rather than raw model access.
What this means is that QueryStory isn’t just about producing AI-generated stories—it’s about proving those stories can be trusted, audited, and costed in ways executives can stand behind.
