Databricks’ new AI agent tells CEO Ali Ghodsi the unvarnished truth
First reported by TechCrunch ·
AI agents now offer business leaders direct, unvarnished truth that bypasses hierarchical filters.
Databricks has unveiled an updated version of its AI agent, Genie, capable of performing multi-step tasks autonomously by leveraging an ontology that provides context from customer data on the Databricks platform. CEO Ali Ghodsi demonstrated the agent's capabilities by showing how he uses it internally for tasks such as analyzing employee satisfaction survey results, evaluating proposed solutions to low scores, summarizing underperforming sales pitches, and preparing for media interviews. Ghodsi highlighted that Genie provides him with unfiltered insights into company issues that might otherwise be hidden from him as CEO, especially in a company of Databricks' size (12,000 employees), where direct communication lines can become diluted. He positions Genie as a source of "unvarnished truth" for business leaders and believes it can offer similar benefits to the business side as AI and agents provide to programmers.
Databricks' enhanced Genie agent signifies a move towards AI copilots that provide executive-level strategic insights, moving beyond task automation to offer critical feedback and analysis that might not reach leadership through traditional channels. This capability is particularly valuable for large organizations where communication can become fragmented, enabling leaders to identify and address systemic issues effectively. The company's emphasis on an "ontology" suggests a sophisticated approach to grounding AI in a company's specific data, promising more relevant and actionable insights than generic AI tools.
The development of tools like Genie aims to mitigate the risks of AI escalating corporate costs, a concern for many businesses. Databricks proposes solutions such as their Unity Gateway to control usage and enable the use of less expensive open-weight models, indicating a market focus on cost-effective AI deployment. This points to a broader trend where AI providers are not only developing powerful agents but also building in mechanisms for governance and economic efficiency to encourage wider enterprise adoption.
AI-written summary. May contain errors.