Salesforce Unveils Agentforce, New AI Interface

While AI models possess vast global knowledge, they typically lack specific context about individual businesses, their customers, inventory, and internal processes. Salesforce positions itself as a crucial partner in providing this missing context.

By connecting diverse data sources within an enterprise, Salesforce creates a foundational 360-degree view, enabling AI to operate with relevant and actionable business intelligence.

This shift redefines the user experience, allowing users to articulate desired outcomes rather than clicking through dashboards or menu options.

The new interface transforms into a dynamic, intelligent, and composable system, reducing the need for humans to perform rote tasks. This paradigm change allows for AI-driven execution and greater efficiency.

Salesforce asserts this marks the end of software that requires humans to handle all the work, opening the door for more intuitive and outcome-focused interactions.

This initiative aimed to unlock "trapped value" embedded in enterprise metadata, which often remains inaccessible within traditional systems.

The Co-Work project provided a complete AI transformation in real-time, showcasing how AI can dynamically adapt and respond to enterprise needs. This pilot served as a foundational step for Salesforce's broader Agentforce strategy.

Marc Benioff defended Salesforce's financial trajectory, countering market perceptions of slowing sales by highlighting significant revenue milestones. He noted the company spent three years reaching $30 billion in revenue, then accelerated to $40 billion in just two years.

Salesforce aims to reach $50 billion in revenue next year, indicating a robust growth path. Benioff's statements underscored the company's continuous expansion and financial strength.

Finn is recognized as the "highest performing customer agent in the world," indicating its advanced capabilities.

This acquisition is a strategic move to bolster Agentforce's ability to deliver superior customer service and automation. The integration enhances Salesforce's comprehensive AI strategy for enterprise clients.

Despite increased AI adoption, many customer applications have shown minimal visual changes over recent years, leading to a disconnect between AI implementation and visible user interface evolution. CIOs and CDOs frequently report a lack of true transformative impact evident in their systems.

Salesforce's new AI-driven interfaces promise to make system evolution visible and accessible to all employees, addressing the current stagnation. This approach aims to provide tangible improvements in how users interact with enterprise software.

Coding, traditionally perceived as a single-player endeavor, is evolving into a collaborative, multiplayer sport with the integration of AI agents. These agents facilitate team-based writing, reviewing, and shipping of code.

This new workflow is seamlessly integrated within Slack, leveraging Slack CRM functionalities to enhance team collaboration. The shift allows developers to work more efficiently and collectively on coding projects.

AI agents streamline the development cycle by enabling shared tasks and real-time feedback within the existing communication infrastructure.

Salesforce is delivering new versions of its platform, including the 'AI Force Max Edition,' which bundles all its AI capabilities into a single core product. This new edition is designed to simplify product acquisition for customers by eliminating uncertainty about required functionalities.

The 'AI Force Max Edition' offers an all-in-one solution for organizations seeking comprehensive AI integration. The goal is to provide a unified product that delivers total success and streamlines deployment.

This plugin tackles issues such as MCP servers and ensuring zero data retention, which are critical for enterprise security.

The solution allows enterprise users to deploy powerful AI capabilities by simply activating the plugin. This significantly reduces the technical barriers to integrating advanced AI into existing systems.

The plugin can be accessed via AppExchange.

To activate the plugin, users need to submit an Org ID. For hands-on assistance with setup, AI Learning Labs are available at Moscone West.

This internal review is crucial for refining current practices and ensuring robustness.

Amodei emphasized the need for increased investment in transparency and safety initiatives across the industry. He advocates for greater openness and dedicated resources to tackle AI-related risks.

He also called for organizing industry-wide standards, suggesting a collaborative approach to establish common benchmarks and best practices. An international component should be incorporated to ensure global consistency in AI safety protocols.

He emphasized the critical need for better diffusion of existing AI technologies across industries.

Amodei highlighted practical use cases such as Sales/Go-to-market analysis, which could greatly benefit from broader AI application. He argued that widespread adoption and integration are necessary to unlock the technology's true economic impact.

However, data grounding remains a critical bottleneck for many firms, impeding their ability to fully leverage AI. Addressing this challenge is essential for realizing the technology's unfulfilled potential.

The current underutilization suggests a significant opportunity for businesses to enhance their operations by more effectively integrating AI solutions.

Marc Benioff used the metaphor of a "generic AI as basically like an amazing race car with absolutely no wheels" to illustrate the limitations of AI without proper context. He asserted that Customer 360 acts as the essential foundation, providing the "wheels and the steering wheel" for AI to function effectively.

This analogy underscores that generic AI models, while powerful, lack the specific business logic and data required for meaningful enterprise application. Salesforce integrates business logic and data as central components, making AI truly actionable for companies.

The Agentic Enterprise concept emphasizes that a holistic view of customer data is indispensable for AI to drive tangible business outcomes.

This model is engineered to support long-running agents that manage complex outcomes within customer relationship management.

COA was trained on synthetic data, extending NVIDIA's NeMoTron framework, to ensure robust performance without compromising customer privacy. A critical aspect of its development is that no actual customer data was used for its training.

This new model enhances the capabilities of Customer 360 by enabling more sophisticated and autonomous decision-making processes.

He posited that this allows enterprises to "know everything and do anything."

Huang highlighted NVIDIA's transformation from a GPU company to a full-stack AI factory, emphasizing the broad impact of AI on business operations. Enterprises are evolving into "agentic enterprises" through this integration.

This shift implies a fundamental change in how businesses operate, leveraging AI to gain comprehensive insights and execute complex tasks.

Huang categorized the safety framework as a computing system problem, emphasizing the need for robust and reliable system design. This perspective shifts the focus towards practical, implementable safety measures.

He highlighted the necessity of developing robust test environments for complex AI systems to ensure their reliability and safety. Proper testing is crucial for identifying and mitigating risks in advanced AI applications.

Open AI models have seen a dramatic increase in market share, growing from 30% to approximately 70% within a year, according to Jensen Huang. During the same period, closed models experienced a 25-fold increase in token generation, indicating overall growth in AI usage.

This surge is largely driven by enterprises increasingly building custom AIs tailored to their specific needs. The shift highlights a trend towards greater flexibility and customization in AI deployment.

Huang suggested that every software company is now, by extension, an AI company, reflecting the pervasive integration of AI across the tech landscape. The shift towards open models and custom builds signifies a maturing AI market where specialization and adaptability are key.

Salesforce introduced Agent Fabric, an evolution of MuleSoft, designed to provide a unified registry for managing all AI agents. This tool simplifies the orchestration of multiple agents across various platforms.

Simultaneously, Salesforce Guardian was unveiled, focusing on agent identity and data classification to ensure secure and compliant AI operations. A key commitment is zero data retention for enterprise intellectual property.

These new tools aim to streamline the management and security of AI agents within an enterprise, offering comprehensive control and protection.

This platform allows for centralized control and oversight of agent deployments.

Simultaneously, Salesforce Guardian was introduced to address the security needs of security leaders, focusing on agent identity, data protection, and compliance. This ensures that AI agents operate within secure frameworks.

The new tools also include features like 'agent fabric wallets' for budget management and automated optimization, providing financial and operational control over AI deployments. These innovations streamline the deployment and maintenance of AI agents.

Denny, CEO of Adecco, reported that their recruiters saved 35-40% of their time thanks to AdzentiKI, their AI system. This significant time saving allowed recruiters to engage in more qualitative conversations with both clients and candidates.

The improved processes, driven by AI, led to Adecco placing an additional 20,000 people year-to-date. The company emphasized that AI must operate "with people and not to people," focusing on augmenting human capabilities.

This practical application demonstrates how AI can enhance productivity and improve the quality of human interactions in the professional services sector.