From the moment the doors opened today, there was a different feeling at Moscone. It wasn’t just about new features — there was a stronger sense that Salesforce is reasserting purpose, reordering its narrative around how all the pieces fit. By lunchtime, the “360” rebrand had already sparked dozens of hallway conversations: “Data Cloud is now Data 360,” “Agentforce is now Agentforce 360.” That might seem cosmetic, but it’s a signal about what Salesforce expects people to focus on going forward: unity, context, and agents that live in the full loop.

The core message: agents, data, people, and action need to flow together. Day 1 was all about showing (not just telling) how far that integration is coming — and how fast.

Major Announcements & Highlights

Here’s what stood out today, beyond the name changes.

1. The 360 Rebrands: What They Mean

• Data Cloud → Data 360
  This reframing positions it less as a backend tool and more as a central nervous system for all AI and automation. The hint is: every agent, every flow, every insight should run through this hub.
• Agentforce → Agentforce 360
  This name change reflects how Salesforce is thinking about agents not as standalone bots, but as continuously context-aware, evolving “teammates” in your processes.

It’s easy to dismiss a rename as marketing. But in this case, the rebrands reinforce Salesforce’s direction: stitching together data, intelligence, and action as one system.

2. Agentforce 360: From Demo to Platform

Salesforce officially launched Agentforce 360 today, updating what was already an ambitious concept into a framework they expect customers to use at scale.

Some highlights:

• It’s not just about chat agents. Salesforce wants Agentforce 360 to host agents that reason, act, monitor, and evolve.
• They introduced Agentforce Commerce, which lets merchants sell via ChatGPT’s Instant Checkout while still controlling fulfillment and data.
• Voice is coming more strongly into view — the agents will be able to speak, detect nuance, and more naturally interact with users.
• Deeper governance, observability, versioning, audit logs — the “Command Center” functions were prominent in demos and product sessions.
• The messaging: this is the year to move past pilots. Salesforce is pushing hard on production maturity, operational scale, and robustness.

If 2024 was about proving agents could do things, 2025 is about proving they should run in the core workflows.

3. Slack as the Agentic User Layer

One of the more narrative-charged moves: Slack is being elevated from collaboration tool to “agentic OS.”

Key aspects:

• Slack will host intelligent assistants that summarize threads, auto-draft responses, extract tasks, and more.
• It will become one of the conversational ports into CRM data, downstream systems, and agents.
• Underneath, there’s a backend — MPC server, integration layer — doing the orchestration between Slack, agents, APIs, and data.
• Over time, Salesforce is showing how someone could never leave Slack to interact with agents or data.

It’s a bold vision: make the interface where people already work the front door to intelligence.

4. AI Partnerships & Multi-Model Strategy

Salesforce announced stronger integrations with both OpenAI’s GPT-5 and Anthropic’s Claude models to power parts of Agentforce 360.

Why this is interesting:

• It signals that Salesforce isn’t tying itself to a single LLM — different models will serve different use cases, verticals, compliance zones.
• Claude is being positioned especially for regulated industries (financial services, healthcare) running in secure environments.
• Embedded access: surprising glimpses of building agents and dashboards inside ChatGPT, bridging the gap between Salesforce and more familiar AI workspaces.

This gives them flexibility and lets them hedge model risk while optimizing per domain.

5. New Verticals & ITSM Inroads

Day 1 also teased some of Salesforce’s ambitions beyond core Sales/Service:

• Agentic IT Service (ITSM) is confirmed to be a new offering. Salesforce aims to compete in the space with a conversational, Slack-embedded approach to incident, request, and service management.
• We heard hints of industry-tailored agents: financial services, healthcare, etc. (That’s consistent with prior releases and rumors heading into Dreamforce.)

These moves show Salesforce is broadening where agentic AI can operate — not just in help desk or customer service, but into internal ops and domain-specific workflows.

6. Reinforced Themes: Governance, Data Trust & Scale

Agentic AI is sexy, but Day 1 made sure to remind us: you cannot skip the infrastructure and guardrails. Some of the quieter but important announcements:

• Salesforce doubled down on data governance, policy-based controls, masking, audit trails, etc. as foundational.
• There were pushbacks against fragmentation: the idea that data scattered across systems kills agent reliability.
• Observability, versioning, rollback, and “safe defaults” were repeatedly emphasized.
• Several speakers talked about “closing the feedback loop” — letting agents learn from mistakes, have guardrails, be auditable.

These are the mechanics that will decide whether agentic AI is enterprise-safe or risky hype.

What This Day Means & What to Watch

At the end of Day 1, here’s what I walked away thinking — and what I’ll be watching in the coming days.

The Tone Has Shifted

Before, Salesforce often felt like it was selling hope — demos of what agents might be able to do. Today, the narrative is more grounded: “Here’s what you can build now, scale now, and operate safely now.” The 360 rebrands help anchor that shift. The expectation is less “future promise” and more “adopt this.”

Execution Will Be the Measure

The real test now is going to be how customers put Agentforce 360 into production — with real SLAs, error budgets, oversight, ROI metrics. Salesforce is giving tools for that (observability, versioning), but the onus will shift to implementers, architects, and governance teams.

Data Is Still (and Will Always Be) the Hard Part

All the AI in the world means little without clean, reliable, integrated data. The Data 360 rebrand reflects that Salesforce knows this is the linchpin. Today’s demos showed how agent outputs choke when context is missing; the safe bet is that a lot of the real “lift” will come from data prep, connectivity, and policy work.

Slack’s Role Is Under the Spotlight

If Slack can really deliver as a conversational layer into CRM and agents, it could simplify adoption paths. But users will quickly feel friction if the backend routing, latency, context switches, or permission edges are messy. Day 1 did a good job showing the promise; the coming sessions need to show stability, resilience, and edge-case handling.

Vertical & Cross-Domain Agents Are the Next Frontier

Customer service is just the beginning. ITSM, finance, operations, marketing — all of these are fertile territory for agents. How well Salesforce scaffolds vertical templates, domain guardrails, and accelerators will shape its ability to compete beyond the “agent toy” zone.