Skip to main content
GPTfy - Salesforce Native AI Platform

Build AI Agents. No Code Required.

GPTfy's Agent Builder lets you create AI agents inside Salesforce in minutes—describe what you want, pick pre-built or custom skills, attach knowledge base, and deploy to Slack, MS Teams, or WhatsApp.

Creating AI agents in Salesforce no longer requires deep development. Describe your use case, choose a template like sales or support, add skills and optional RAG, then expose the agent on the channels where your team already works.

What's inside this demo

Describe and Generate

  • Choose a button (e.g. Sales, Support, Help Desk) or describe in your own words what you want the agent to do.
  • A system prompt and agent description are generated automatically; you can test the agent before publishing.

Skills and Knowledge Base

  • Attach skills such as create case, update opportunity, create account; add or build custom skills for your use cases.
  • Optionally give the agent access to your GPTfy knowledge base (RAG) so it can resolve cases using your content.

Channels and Security

  • Deploy the agent to Slack, MS Teams, WhatsApp, Google Assistant, or MS Copilot so users work where they already are.
  • Control access via permission sets so different roles (e.g. sales rep vs admin) get appropriate capabilities.

Use this video when

You need a sales or support agent in Salesforce and want to avoid custom code.

You want to expose CRM actions (e.g. create quote, update opportunity) via Slack or Teams.

You need a help-desk or case-resolution agent with access to your knowledge base.

You want to bring your own AI model and still use GPTfy for agent orchestration and channels.

Frequently asked questions

The Agent Builder demo walks through creating a fully functional AI agent inside Salesforce in under five minutes using GPTfy. You start by choosing a pre-configured template (Sales, Support, or Help Desk) or by typing a plain-language description of what you want the agent to do. GPTfy automatically generates the system prompt and agent description, then lets you test the agent before publishing. The demo covers attaching skills (e.g. create case, update opportunity), connecting a knowledge base for RAG-based resolution, selecting an AI model via Bring Your Own Model (BYOM), deploying to channels like Slack, MS Teams, or WhatsApp, and configuring per-role permission sets.

GPTfy Agent Builder is a Salesforce-native managed package and works with any Salesforce org that supports Apex and Lightning, including Sales Cloud, Service Cloud, Experience Cloud, and CPQ editions. It is compatible with Enterprise and Unlimited editions and can be configured in Developer orgs for testing. Because it is a managed package installed from AppExchange, no custom metadata deployment is needed beyond the configuration steps shown in the demo.

After the agent is generated, you navigate to the Skills section where you see available skills (e.g. Create Account, Create Case, Update Opportunity) and selected skills. You drag or click to move skills into the agent's selected set—no code required. For the knowledge base, you select one or more RAG data sources (GPTfy RAG Sync collections) that you want the agent to reference when answering questions. For example, connecting a support knowledge base lets the agent resolve cases by looking up relevant articles automatically. Custom skills can also be built for specific use cases.

Security is controlled through Salesforce Permission Sets. Admins assign permission sets to users or roles, which determines what the agent can access and what actions it can take. For example, a sales rep permission set might allow the agent to create quotes and update opportunities, while a service rep set allows case creation and resolution. No data leaves your Salesforce org unless you explicitly configure a named external AI provider. The BYOM (Bring Your Own Model) capability lets you use a model already approved by your organization—so you stay within existing enterprise data-governance boundaries.

Agent Builder is designed to meet users where they already work. The channels feature lets you deploy the agent to Slack, MS Teams, WhatsApp, Google Assistant, or MS Copilot—so your team interacts with Salesforce data through the tools they already use daily. On the Salesforce side, the agent uses skills to perform real CRM actions (create cases, update records, search accounts) so it becomes part of the workflow rather than a separate tool. You can also configure a welcome message and brand the agent with a custom name so it feels native to your organization.

Agent Builder is best when you need to get an AI agent to production quickly without writing Apex, and when your use case maps well to standard Salesforce actions (cases, opportunities, accounts, quotes). It is particularly valuable for teams that need to deploy to multiple channels (Slack and Teams simultaneously) or want to iterate quickly on agent behavior by editing the system prompt. A custom Apex chatbot is more appropriate when you need deeply custom business logic, non-standard data flows, or integrations with systems that GPTfy skills do not yet cover. For most enterprise use cases—sales qualification, case resolution, meeting prep—Agent Builder provides 80–90% of the capability at a fraction of the implementation time.

Ready to see this in your Salesforce org?

Book a 45-minute session and we'll walk through this use case using your own data.

Book a Demo
Video transcript
Hello everyone, and I have something really interesting that I want to share with you guys. So in this demo, we are gonna explore how GPTfy enables you to create agents inside Salesforce. We have tried really hard to make this process as easy as possible for you guys. So here I am inside Salesforce, and this is the GPTfy corporate. Everything around GPTfy runs from here and we have agents, prompt catalogs, skills and prompts, knowledge base and data text mapping. So a lot of things. But we are gonna focus on building a dev and agent so you can see how it goes. So this is the agent builder that we have created, and it is as easy as choosing a button. Writing, just describing what you want your agent to do, right? Just describe it. And we have created some, most common use cases that we get from our customers, like they want a sales agent, want a support agent or a help desk, right? You can either choose to build me an agent, which can help me write. Resolve cases, right? Or I can just simply choose to click the sales button. And a pre-configured prompt just populates, right? Create a sales agent that helps manage tasks, track opportunities, qualify prospects, provide product information, generate codes, answer pricing questions, and guide customers through the sales process. This is a pre-configured prompt and I'm just gonna hit send. Alright. And a completely new agent will be generated right in just a second. Success. So this is a completely new agent that GPTfy created, and it has given it an agent description. It has created a system prompt all on its own based on what description we gave it, right? GPTfy enables you to test out the agent before you directly publish it, and you can test the agent over here in this text box, right? And now you can add multiple skills to this particular agent. So you know, you can see the available skills and the selected skills in the available skills section. If we have create account on closure case, create account name, check, create a case, create case for contact, multiple skills that you can associate or give to a particular agent based on what you want that particular agent to do for you. And we have enabled the user, they can develop new skills according to their particular use cases as well. Update opportunity and it's as easy as selecting it, adding it over here. You see we just moved update opportunity over here and now this particular agent has that particular skill. Pretty straightforward. Moving on. To the AI model. We have an agent model for this particular demo environment and you guys can choose as GPTfy as a bring your own model. We can configure or we can leverage any model that your company or enterprise has already invested in. We moving on, we also have the ability to provide the agent to access a company's knowledge base. So for example, I want the new cases that are getting created and I want this particular agent to resolve this using my GPTfy agent, right? So I can just select it, move it over there, and now this particular agent has the access to the GPTfy rack. As simple as that very, non code and no code type fee to it. And the best part about it, we have a section, as you can see here, called channels. And what that does is basically it enables your team where they already work, right? Most of the teams work. MS Team, or they're using MS Copilot, they're managing their day to day through Slack or right. Google Assistant WhatsApp. And you can just enable this particular agent to any of this, these channels, so you can just access them right to here. So for WhatsApp for example, this agent, I can enable it and I can just start texting it via WhatsApp and. That can, and that's how convenient GPTfy has made this whole process. You can just basically ask your particular agent to give you like your Salesforce information from your org to MS teams or WhatsApp or copilot, wherever you want it. And there are various security protocols set over here. You can, based on users, you can assign them permission sets and that gives the admin, complete control of who is accessing what, what like a sales rep can ask a certain type of questions and it'll be different for a system admin or a sales manager. So you, everything around that can be controlled from here. We also have an advanced. Setting section, which basically has, the welcome message, which over here that can also be configured. You can add your company's name or like you must have seen, a lot of companies have they want to personalize it. So they would give a name to their agent or their board that is talking. So it can say Hey, this is GPTfy bot, or some name, let's say WALL-E for example. Or, hi, this is WALL-E and you can configure everything. Like that from there. And this is it then. This is how simple and straightforward GPTfy has made the creation of agents inside Salesforce. We at GPTfy believe that it shouldn't be that hard to leverage or implement AI inside Salesforce. And if you like this particular demo do you want to see more of this in action or you would like to see maybe your, any particular use cases, please do book a call with us and we would love to have a conversation with you. Thank you so much.

Explore More

Demo Videos

Browse all Salesforce and AI demo videos.

AI for Sales

See how GPTfy accelerates the full sales cycle.

CPQ Quote Creation (Slack)

Create CPQ quotes from Slack with the GPTfy agent.

Last updated: March 2026