TLDRs;
- Salesforce unveiled Agentforce 360 at Dreamforce 2025 to connect humans and AI across business workflows.
- Over 12,000 companies, including Reddit and Adecco, are already using the platform to boost efficiency.
- Complex pricing and credit-based billing could slow adoption among large enterprises.
- Salesforce’s partnerships with Databricks and Snowflake aim to fix data challenges critical for enterprise AI success.
Salesforce has officially unveiled Agentforce 360, a groundbreaking platform designed to bridge human employees and AI agents across all core business operations.
The announcement took place at Dreamforce 2025 in San Francisco, marking one of the company’s most ambitious steps yet into the enterprise AI space.
The new platform brings together Salesforce’s existing ecosystem, spanning sales, service, advertising, and IT, under a unified AI layer. Integrated natively with Slack, Agentforce 360 empowers companies to automate repetitive workflows, streamline communication, and enhance decision-making across teams.
Developed over the past year with direct input from thousands of enterprise users, Agentforce 360 has already attracted over 12,000 customers, including major names like Reddit, Adecco, and OpenTable. Salesforce claims the platform has delivered measurable efficiency gains: Reddit reportedly deflected 46% of its customer support cases, while Engine cut handle time by 15% using Agentforce-powered automations.
Early Results Show Real-World Impact
Salesforce says the platform is more than a set of automation tools, it’s an intelligent framework for human-AI collaboration. By combining natural language interfaces, predictive analytics, and adaptive workflows, Agentforce 360 enables organizations to create “digital teammates” that learn from user behavior and continuously refine their output.
“The goal is not to replace workers,” a Salesforce spokesperson said during the launch. “It’s to give every employee a team of AI agents that make their work faster, smarter, and more meaningful.”
The rollout reflects Salesforce’s broader vision for AI-first enterprises, where every digital process, from marketing campaigns to support ticket triage, operates in tandem with AI-driven insight.
Salesforce just dropped Agentforce 360 — its biggest push yet to make AI actually useful inside enterprises. The goal? Help companies escape “AI pilot purgatory” and turn prototypes into production-ready agents. pic.twitter.com/CBt9orlAuf
— Ken Yeung (@thekenyeung) October 13, 2025
Pricing Sparks Debate Among Enterprises
While Agentforce 360’s capabilities have been well received, its pricing model has drawn scrutiny from industry analysts. Salesforce offers two main options, a conversation-based plan at $2 per interaction, or Flex Credits priced at $0.10 per action (equal to 20 credits). Customers must choose one model per Salesforce organization, meaning they can’t mix pricing strategies within the same tenant.
That restriction forces businesses to estimate usage before they have real data, a challenge for those piloting AI integrations at scale. Salesforce does allow switching between plans by changing SKUs, but the process introduces extra friction for finance and operations teams.
To address cost transparency, Salesforce built an ROI Estimator and Digital Wallet feature, an in-app ledger that tracks usage and spending in real time. Still, many teams lack baseline metrics for how many AI “actions” their processes consume, making accurate budgeting difficult. Analysts warn that miscalculations could inflate costs, potentially slowing adoption despite the platform’s promise.
Demand Grows for Data and Semantic Layers
Beyond pricing, experts say Agentforce 360 highlights another challenge,data readiness. For AI agents to function optimally, they require structured, contextual data, something many enterprises still lack. Salesforce is responding by integrating Tableau Semantics with Data Cloud 360, a feature that unifies data definitions across departments.
To strengthen this layer, Salesforce is partnering with Databricks, dbt Labs, and Snowflake to create standardized “semantic vocabularies” that AI agents can easily interpret. According to Gartner, up to 60% of AI projects could fail by 2026 without solid data foundations. That creates new opportunities for third-party vendors specializing in data preparation, semantic mapping, and document processing acceleration.
Firms capable of rapidly structuring unorganized data, respecting permission layers, and ensuring contextual accuracy are expected to see rising demand as companies adopt Agentforce 360.
