AI's trillion-dollar opportunity: Context graphs (foundationcapital.com)

A new analysis highlights the emergence of context graphs as a transformative opportunity within the AI/ML space, arguing they could pave the way for a new generation of decision-making systems. Jamin Ball's post, "Long Live Systems of Record," explores how traditional systems like Salesforce, Workday, and SAP may evolve rather than be replaced by AI agents, which act across multiple datasets and workflows. The real challenge is not in data retrieval but in building systems that capture 'decision traces'—recorded rationale and exceptions that accompany decisions. These traces, currently lost in conversations or informal communications, are essential for understanding the context behind actions taken by AI agents during operations. Context graphs emerge as a solution, providing a structured and queryable framework of decision lineage that connects historical decisions to contemporary actions. As agents interact with various data from multiple systems in real-time, they generate this crucial context, detailing how decisions were made and the reasoning behind them. By successfully integrating these decision traces, startups in the AI space can create robust systems that not only automate tasks but also offer a historical perspective on decision-making processes, turning exceptions into searchable precedents. This positions context graphs as pivotal assets that could redefine enterprise operations in the burgeoning era of AI.