Salesforce's head of data architecture says knowledge graphs are becoming critical infrastructure for enterprise AI systems.
As organizations move toward AI-native IT stacks, they are building layered data architectures that sit above fragmented data sources—lakehouses, operational databases, and customer profile stores—to give AI models the context needed to reason reliably.
The core challenge is not technology but governance: semantic descriptions of data and access policies must move up the stack so that AI can return trustworthy, explainable answers at conversational speed.
What happened
Salesforce's head of data architecture, Tristan Baker, outlined how knowledge graphs are becoming a foundational layer for enterprise AI systems. He described a layered stack where lakehouses and operational databases sit below a metadata layer, with graph technology managing relationships and context across them—enabling AI agents to reason from trusted data rather than simply retrieve it.
Why it matters
Enterprise leaders are seeking conversational interfaces that return trustworthy answers in seconds, but that requires far more than a single database. Baker emphasized that the hard problem is accurately describing enterprise information so AI can make sense of it across fragmented data landscapes. Governance and semantic access policies are emerging as critical challenges; legal teams need to define data protection at the semantic level, not just within individual databases.
What to watch
Baker noted that most organizations have not yet fully solved this architecture problem. He cautioned that even the best technology yields "garbage in, garbage out" if the content exposed to it is not carefully curated—a challenge many enterprises are still grappling with.
Ask the AI about this article →
Enterprise organizations are shifting from cloud-native IT stacks to AI-native ones, and that transition is reshaping how they think about data architecture. The move reflects a fundamental change in what enterprises expect from data systems: no longer just speed and scale, but the ability to deliver trustworthy, explainable answers in a conversational interface. Knowledge graphs have emerged as the connective tissue binding fragmented data sources together—lakehouses, operational databases, customer profile stores—into a coherent system of intelligence.
Baker's framing reveals why this shift is non-trivial. The problem is not that enterprises lack data; it is that their data lives scattered across dozens of copies in incompatible systems, with no shared semantic understanding of what it means. A metadata layer can bridge that gap by translating business language into database columns, and graph technology can track relationships and context across those translations. But without careful governance—semantic policies that legal and security teams can understand and enforce—even the best AI reasoning engine will produce unreliable answers. Most organizations, Baker suggests, are still struggling to design these governance layers at the semantic level rather than hiding them inside individual databases.
For example, today's edition would include:
AI-summarized, only the topics you pick — one digest a day via Email, Slack, or Discord.
Free · takes 30 seconds · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. Q&As are published on this page for other readers too.
Phonely Ltd. launched Alma, a large language AI model built for voice agents and trained on over 10 million re…
Aranya Inc., a startup founded last year, launched today with $11 million in funding
CBTS Technology Solutions LLC launched Forge Agents, a platform that turns a plain-language job description in…
Imec CEO Patrick Vandenameele said at SEMICON Taiwan 2026 that the Belgian research center is broadening its c…
Alphabet's AI Overviews now reach over 2.5 billion monthly users through Google Search, and its ad business ge…
Sarah O’Connor's book 'We Are Not Machines' explores how mechanization and AI have transformed the workforce…
