Hi Trailblazers.,
I’m exploring Agentforce
and would love to learn from teams who have already implemented it.
If you’ve deployed Agentforce in any capacity, could you share your experience?
Specifically, I’d love to know:
1. What did you use Agentforce for?
- What business problem were you trying to solve?
- Was it customer-facing, agent-facing, or used for back-office automation?
- Which parts of your Salesforce stack did you integrate with (Service Cloud, Data Cloud, Omni-Channel, Knowledge, etc.)?
2. How did you deploy it?
- Architecture or setup approach
- Custom prompts vs. out-of-the-box configurations
- How you handled grounding, data governance, and security
- How you integrated it into existing workflows (e.g., case swarming, chat, voice, digital channels)
3. Did it work?
- What outcomes did you see (qualitative or quantitative)?
- What worked exceptionally well?
- What didn’t work or required rethinking?
- Any surprises during rollout?
4. What would you ask or consider before deploying again?
Any lessons learned around:
- Change management
- Agent adoption
- Hallucinations or accuracy
- Scalability
- Required cleanup of knowledge articles/data
- Measuring success
- Skills or personas required to maintain Agentforce long-term
5. Any clever tips or things you wish you knew earlier?
- Configuration tricks
- Testing strategies
- Guardrails or safety measures
- Prompt engineering best practices
- Organizational prerequisites
Thanks in advance for sharing your insights! Your real-world experiences will really help guide our approach.
#Agentforce
As a company that implements Salesforce and Agentforce, we'd be happy to share our experience.
One thing we've learned is that it's better to start with one specific business problem rather than trying to automate everything at once. Customer support is often a good place to begin. For example, Agentforce can handle routine inquiries, find relevant information, and assist human agents.
Good data is another big part of the equation. If your knowledge articles are outdated or your CRM data is messy, the agent won't be able to deliver reliable answers.
We'd also recommend involving the people who will actually use the solution. Testing real scenarios and collecting feedback after launch are just as important as the initial setup.