Salesforce announced a new portfolio of job-ready Agentforce agents on September 11, 2026, timed to its Dreamforce 2026 conference (September 15–17 in San Francisco). The headline shift: instead of starting with a blank agent-building platform, companies can now start with seven named, prebuilt agents designed around specific jobs in service, sales, commerce, IT and HR, and the back office.
According to Salesforce, each agent ships with the skills, actions, and data models required for its job, connects to Customer 360 context, and can be tailored — including a custom name — to each business. Six of the seven are generally available now. This post breaks down what was announced, who it affects, and how to evaluate the portfolio against your real processes before buying.
Agentforce job-ready agents are Salesforce's new portfolio of seven prebuilt, named AI agents — Casey, Paige, Carter, Hunter, Marshall, Piper, and Fin — each scoped to a specific business function and announced September 11, 2026 alongside Dreamforce 2026. This matters most for Salesforce customers in service, sales, commerce, IT/HR, and operations who want faster time-to-value than a build-from-scratch agent project. The decision this article supports: which agents map to processes you actually run, and whether your data, permissions, and process documentation are ready. Vantage Point helps teams make exactly that evaluation as part of our Salesforce implementation and advisory work.
Salesforce's announcement names seven agents, each built for a specific job. Six are GA now; one is in pilot. Here is the portfolio as Salesforce describes it:
| Agent | Job | Scope (per Salesforce) | Availability |
|---|---|---|---|
| Casey | Customer service help agent | Resolves issues across voice, SMS, WhatsApp, and web chat; pre-built support for FAQs, returns, account management, and human escalation | GA now |
| Paige | IT & HR service agent | Resolves employee requests across Slack, portals, and the tools employees already use | GA now |
| Carter | Shopper agent | Helps shoppers discover and compare products, get answers, and convert with in-chat checkout | GA now |
| Hunter | Outbound sales agent | Works a pipeline from research to outreach, collaborating with sellers over weeks and months | Pilot now; GA November 2026 |
| Marshall | Supply chain agent | Orchestrates end-to-end back-office processes, automates manual work with deterministic execution, and keeps an audit record of every action | GA now |
| Piper | Inbound pipeline generation agent | Works across websites and inboxes to engage, qualify, and convert inbound leads for B2B sales and marketing teams | GA now |
| Fin | Customer agent | Resolves complex customer experience workflows across every channel, powered by Operator and Fin Apex custom models | GA now |
Two things stand out in the details. First, this is not a fully homegrown list: Salesforce's own announcement links Piper to Qualified and Fin to fin.ai, so parts of the portfolio arrive through partner technology rather than net-new Salesforce engineering. Second, Salesforce backs the launch with early customer metrics it attributes to these agents — for example, it says 50% of Engine's chat inquiries are fully resolved by its help agent, 60% of Perk's sales pipeline is built by Hunter, and 79% of the conversations Fin sees at Anthropic are resolved autonomously. Treat those as vendor-reported results, not guarantees.
For a broader view of how these agents fit the rest of the Agentforce product line, see our guide to the Salesforce Agentforce ecosystem.
The agents are the headline, but the enabling capabilities matter as much for planning:
This announcement touches any organization already on Salesforce — or evaluating it — in five functions:
Admins and platform owners are affected in every scenario: each agent still needs configuration, data access decisions, and guardrails. Teams on HubSpot or mixed CRM stacks are not the direct audience, but the "prebuilt workforce" pattern is worth watching — expect comparable packaged-agent offerings across the CRM ecosystem.
The risk with a prebuilt portfolio is buying a name instead of solving a process. Before committing, work through four steps:
Use this readiness checklist as a starting point:
| Readiness area | Question to answer | Red flag |
|---|---|---|
| Data | Is the customer/employee data the agent needs clean, unified, and accessible in Salesforce? | Duplicate records, missing fields, data stuck in silos |
| Permissions | Can you scope exactly what the agent may see and do, by profile and object? | "Give it admin access to start" |
| Process documentation | Is the target process written down, including exceptions? | The process lives in one person's head |
| Escalation design | Is there a defined handoff to humans, with context preserved? | Escalation means "start the conversation over" |
| Measurement | Do you have baseline metrics (handle time, resolution rate, pipeline created)? | No baseline means no way to prove value |
| Change management | Do the people working alongside the agent know what changes for them? | Agents launched at teams, not with them |
If several red flags apply, close those gaps first — workflow automation and process optimization work often has to precede agent deployment, and advisory and change management support determines whether teams actually adopt what you deploy.
| Date | What happens |
|---|---|
| Now (September 2026) | GA: Casey, Paige, Carter, Marshall, Piper, Fin, and Multi-Agent Orchestration. Pilot: Hunter, Agentforce Coworker AI Skills |
| October 2026 | GA: Agentforce Coworker AI Skills and Agent Optimizer (per Salesforce) |
| November 2026 | GA: Hunter, the outbound sales agent (per Salesforce) |
One caution straight from Salesforce's own announcement: the release notes that some referenced services or features are still in development, and that customers should make purchase decisions based on currently available features. Plan around GA dates, not demo dates.
Vantage Point is a boutique, senior-led Salesforce and HubSpot consulting partner. For teams evaluating the job-ready Agentforce portfolio, we typically help in three ways:
If your team is evaluating how the new Agentforce portfolio applies to your service, sales, or operations processes, talk to Vantage Point about a readiness assessment and a practical implementation plan.
They are a portfolio of seven prebuilt, named AI agents Salesforce announced on September 11, 2026: Casey (customer service), Paige (IT and HR service), Carter (shopping), Hunter (outbound sales), Marshall (supply chain), Piper (inbound pipeline generation), and Fin (customer experience). Each ships with the skills, actions, and data models for a specific job, and each can be tailored and renamed to fit your business.
According to Salesforce, six are GA as of the announcement: Casey, Paige, Carter, Marshall, Piper, and Fin. Hunter, the outbound sales agent, is in pilot now with GA planned for November 2026. Multi-Agent Orchestration is also GA, while Agentforce Coworker AI Skills and Agent Optimizer reach GA in October 2026.
No. Job-ready agents give you a faster starting point for common functions, but they still require configuration, data access decisions, guardrails, and tailoring to your processes. Custom Agentforce development remains the right path for processes that are unique to your business or span systems outside the prebuilt agents' scope.
Four things matter most: clean, accessible data in Salesforce; a permission model that scopes what the agent can see and do; written process documentation including exceptions and escalation paths; and baseline metrics so you can measure impact. Salesforce's own materials emphasize that agents operate within your business rules, permissions, and security — so those need to be right first.
Salesforce's September 11 announcement does not publish portfolio pricing. Agentforce pricing has historically varied by product and consumption model, so confirm current packaging and pricing directly with Salesforce or your account executive before budgeting — and base purchase decisions on features that are currently available.
It is a new Agentforce runtime that lets agents pursue goals over days and weeks rather than completing single interactions. Salesforce describes three underlying capabilities: memory that preserves context across sessions, durable execution that keeps plans running and allows course correction, and dynamic steering that adapts to user feedback. Hunter is the first agent built on it.
Directly, it doesn't — these agents run on Salesforce and Customer 360. Indirectly, it signals where the CRM market is heading: prebuilt, job-specific agents instead of blank-slate builders. If you run HubSpot alongside Salesforce, watch for how agent work crosses both systems, and design your integration and data model so either platform's agents can act on accurate data. Vantage Point works across both platforms and can help you plan that coexistence.