Frequently Asked Questions

Agentforce is Salesforce's AI agent platform that lets you build autonomous agents that can qualify leads, resolve support tickets, book meetings, and execute workflows — without human intervention. Unlike chatbots, these agents can reason, plan, and take action across your Salesforce data.

Einstein AI provides predictions and recommendations — it tells you what might happen. Agentforce agents take action — they actually do the work. Think of Einstein as an advisor and Agentforce as an employee.

Not necessarily, but it helps. Data Cloud gives your agents more context by unifying customer data from multiple sources. For simple use cases, you can start with just your Salesforce data. For more sophisticated agents, Data Cloud makes them smarter.

We specialize in small and mid-size businesses — typically $10M to $50M in annual revenue. These companies are big enough to benefit from AI but small enough that we can make a real impact.

Primarily yes. Our sweet spot is companies already running Salesforce who want to get more value from their investment. We also do new implementations for companies planning to adopt Salesforce.

Absolutely. We build custom AI agents, workflow automation, and intelligent integrations using tools like Claude, OpenAI, Gemini, and open-source models. If you have a business process that could benefit from AI, we can help.

A basic agent can be live in 2-4 weeks. More complex implementations with multiple agents, custom actions, and integrations typically take 6-12 weeks. We'll give you a realistic timeline after our intro call.

It depends on scope. A simple agent implementation would start at $5k. We'll provide a detailed quote after understanding your goals.

Your data never leaves Salesforce. Agentforce runs natively on the platform, which means it reads and writes to your existing objects using your existing security model. There are no external databases, no third-party copies, and no shadow storage happening behind the scenes. The AI processes everything in real time and respects every permission, sharing rule, and field-level security setting you already have configured in your org.

Every agent we build ships with guardrails baked in from day one. We define topic boundaries so the agent knows exactly what it can and cannot discuss, set explicit action permissions so it only does what you've approved, configure escalation rules so it knows when to hand off to a human, and enable audit logging so every decision the agent makes is fully traceable. On top of that, Salesforce's Trust Layer automatically masks sensitive data before it ever reaches an LLM, blocks prompt injection attempts, and enforces your compliance policies. We configure all of this as a standard part of every implementation — it's not an add-on.

No — and this is one of the biggest advantages of building on Salesforce. Agentforce inherits your org's row-level security, sharing rules, and field-level permissions automatically. That means an agent running in a customer-facing context can only see what that customer's profile is allowed to see. An agent running internally respects the same role hierarchy your human users follow. Before we go live, we validate all of this by testing with multiple user profiles to make sure the boundaries hold.

Salesforce maintains SOC 2 Type II, ISO 27001, GDPR compliance, and HIPAA eligibility across its entire platform — and Agentforce inherits all of those certifications because it runs natively on that same infrastructure. If your Salesforce org is already compliant, your agents will be too. For clients in regulated industries, we also configure additional controls like data masking, consent tracking, retention policies, and detailed audit trails that satisfy most compliance review boards out of the box.

You get full transparency into everything. Every agent action is logged — what it decided, what data it accessed, what actions it took, and why it made those choices. You can review all of this through standard Salesforce reports and dashboards, which means your existing team already knows how to use it. For Agentforce specifically, the Atlas Reasoning Engine logs its complete chain-of-thought reasoning, so you can see not just what happened but the exact logic behind each decision. No black boxes.

Salesforce charges per conversation for Agentforce agents, where a conversation means one complete interaction — not per message or per API call. For most small and mid-size businesses, this ends up being significantly cheaper than the equivalent human labor, especially for high-volume use cases like support ticket triage or lead qualification. During the intro call, we'll help you estimate costs based on your expected conversation volume, and we build monitoring dashboards into every implementation so you can track spend in real time. No surprise bills at the end of the month.

Good question — and one we help answer in our intro call. Generally, you need clean data, well-defined processes, and a clear use case. If your org is a mess, we might recommend optimization before AI.

Book a 30-minute strategy session. We'll discuss your goals, review your current Salesforce setup, and identify whether AI makes sense for your situation. No pitch, no pressure — just an honest conversation.

Yes. AI agents need tuning and optimization over time. We offer monthly retainers for ongoing support, monitoring, and capability expansion. Most clients stay with us after their initial implementation.