**London**: Peter Bailis from Google Cloud discusses the pivotal role of generative AI in supplier relationship management, highlighting how AI tools enhance data analytics accessibility while ensuring privacy, reliability, and user engagement in decision-making processes across various sectors.
In a recent discussion with Peter Bailis, Vice President of Engineering at Google Cloud, insights were shared about the emerging role of generative AI in transforming supplier relationship management (SRM). Bailis, whose background includes founding Sisu Data and a tenure at Stanford University, now focuses on leveraging AI within the data analytics framework of Google Cloud, particularly through its proprietary tools like Gemini and Looker.
During their conversation, Bailis emphasised the importance of conversational agents in making data analytics more accessible for users. “I joined Google about a year ago to lead generative AI for our data analytics stack,” said Bailis. “We have, perhaps uniquely among cloud hyperscalers, our own core foundation model family with Gemini.” This development allows Google Cloud to utilise large quantities of user and company data while ensuring that the data remains private and secure for clients.
The traditional method of obtaining insights typically involves navigating various applications. However, with a focus on convenience, Google Cloud integrates AI capabilities into everyday collaboration tools. This strategy enables users to engage directly with data without needing to switch between different applications or platforms. The generative AI is designed to build trust with users, ensuring they understand how outputs are generated and that these outputs maintain their relevance as data evolves.
Illustrating this approach, Bailis provided an example where users query an AI agent about call reasons in a hypothetical call center setup. The agent, adeptly interfacing with BigQuery and employing Looker’s analytic tools, can produce outputs resembling those from human analysts. For instance, a user can ask for the breakdown of call reasons, which triggers the AI to pull relevant data, conduct analyses, and produce visual breakdowns seamlessly all within a chat interface.
Bailis distinguishes the current functionality of generative AI agents from earlier chatbot models, stating, “In the early days of Gen AI, I had a chatbot or a large language model trained on data… Here the LLM is acting on my behalf.” He further adds that these AI agents autonomously conduct multi-step operations, allowing users to access complex insights, such as projected call volumes.
Concerns regarding AI accuracy were also addressed. Bailis assured that Google’s architecture offers high reliability when retrieving answers. He explained that by connecting the AI agent to established data definitions within Looker, the responses can be trusted against a backdrop of verified metrics, reducing instances of misinformation typical in earlier AI models.
Bailis discussed practical applications within various sectors. For example, a telecommunications company is exploring the implementation of conversational analytics agents for its retail associates to enhance customer service interaction by providing immediate data at their fingertips. Here, the AI agent is not only required to deliver precise answers but also gauge user proficiency to ensure appropriateness of responses.
Additionally, Google Cloud’s AI tools tackle different data types, showcasing versatility across structured, semi-structured, and unstructured data. Users can even incorporate images into their interactions, whereby the AI assesses visual inputs alongside textual queries to deliver comprehensive answers, further bridging the gap between user needs and data accessibility.
The conversation also delved into Google’s approach to enhancing user engagement with data, highlighting a shift from static report generation to interactive, real-time analysis. As companies seek to derive actionable insights from vast data sets, an integrated agent capable of notifying users of changes in metrics is seen as the next advancement in the analytic landscape.
Despite this focus on Google’s products, Bailis underscored a broader vision: an open approach to AI integration across various platforms. He stated that their strategy prioritises offering APIs to facilitate implementations across different environments, fostering interoperability between various AI frameworks, including those from competitors. This openness is seen as a pathway to leverage collective data residing in cloud environments, thereby enriching supplier relationships through more informed decision-making.
In conclusion, as organisations increasingly rely on data for strategic decisions, the integration of generative AI within SRM practices holds significant potential to enhance supplier interactions, streamline data queries, and ultimately contribute to more robust supplier relationships. This aligns with a growing market demand for intelligent business solutions that not only deliver insights but also empower users to make informed decisions within their existing workflows.
Source: Noah Wire Services
Subscribe to Industry Updates
