Hello everyone, recently I have been thinking, and also working with Agentic AI, I'm a RevOps Engineer, I'm interested in the implementation of agents to RevOps, but there are some constraints regarding human-in-the-loop, HITL, and security. If you're working with a similar stack, please let's exchange ideas...
What's a RevOps Engineer?
A RevOps Engineer builds automated pipelines and systems to connect sales, marketing, and revenue data seamlessly, eliminating manual work so leads move through the business automatically...
Hey Bright — HITL and security are exactly where Agentforce gets tricky. We're Clouddora — Salesforce consultancy working with AI agents in RevOps. Would be great to exchange notes — what stack are you using?
Thanks Valiantsina H., it's exactly the constraints preventing most agencies from evolving, I use Google cloud run, vultr, vertex AI, and n8n... I will check out clouddora.
How is that different than what RevOps has been doing?
It's more like an advanced feature for RevOps...
Traditional RevOps mostly focuses on linear workflow rules, static CRM hygiene, and manual oversight. A RevOps Engineer designs autonomous, event-driven infrastructure—handling complex real-time decisioning, multi-step AI reasoning, and API orchestrations (like background lead qualification or voice triggers) while enforcing strict security and governance parameters before data ever hits the CRM.
Bright, you might not need that many tools at this point. I retired n8n a few months ago, there are MCP connections for almost every tool nowadays, including Salesforce.
Lolita T. MCP is great for letting AI talk directly to tools like Salesforce, but it doesn't handle background jobs, webhooks, or retries when an AI isn't actively running.. You still need an engine like n8n for background automation, they actually work best together... Curious though...how are you triggering those MCP connections on asynchronous background events without an orchestration layer?
