The ultimate guide to Salesforce AI Knowledge Creation

It's a familiar feeling for any support team: you spend your day answering the same questions over and over. You know the answer is somewhere, but it's not written down, so your team is stuck reinventing the wheel every week. This is a huge time sink, and getting agents to manually write knowledge base articles after a long day is a tough ask.

Salesforce's answer is a tool called Salesforce AI Knowledge Creation. It uses AI to draft knowledge base articles directly from customer chats and emails. This guide will walk you through what it is, how to set it up, what it can do, and, most importantly, where it falls short. By the end, you'll have a clear idea if it’s the right tool for your team or if you’d be better off with something more flexible.

What is Salesforce AI Knowledge Creation?

Salesforce AI Knowledge Creation is an AI feature built into the Service Cloud that’s meant to help agents create knowledge articles without all the usual friction. Instead of starting from a blank page after solving a tough case, agents can let the AI do the initial legwork.

Here’s the gist of how it works: the AI reads through customer conversations from your support channels (think chat, email, etc.). From that transcript, it automatically generates a draft article, complete with a title and sections for the problem, cause, and solution. The agent just has to look it over, make any necessary edits, and hit publish.

The idea is to capture useful information as it happens, making it easier to build a knowledge base that actually helps people. It’s one of the many AI tools under the Salesforce Einstein umbrella, which is their larger effort to add AI features across their platform.

How to set up Salesforce AI Knowledge Creation

Getting this feature running isn't as simple as flipping a switch. It takes a bit of admin work and, critically, you have to be on the right Salesforce plan to even have the option.

First, let's talk about the requirements. This feature isn't available to everyone. You’ll generally need a higher-end plan like the Unlimited Edition, and you'll need to have both Lightning Knowledge and Einstein Generative AI turned on. For a lot of teams, that price jump can be a dealbreaker right from the start.

If you do have the correct licenses, here’s a quick look at the setup process:

1. Turn the feature on: An admin has to go into the Einstein setup page in Salesforce and enable "Einstein Knowledge Creation."

2. Map the fields: This is the most technical part. You have to tell the AI where to put the information it pulls from a conversation. For instance, you'll need to map the AI’s "Problem" output to the "Problem" field in your knowledge article template. This means you already need a well-organized knowledge base with distinct fields.

3. Grant access to your team: The last step is to give the "Knowledge Creation" permission to all the agents who will be using it. You have to assign this to each user, which can be tedious.

If that sounds like a lot of clicking around in settings, you're right. This kind of multi-step, admin-heavy setup is pretty common for enterprise software, but it’s not the only way to do things. Modern tools like eesel AI are built for simplicity, offering one-click integrations that can get you up and running in a few minutes. Instead of wrestling with permission sets, you can connect your help desk and have a working AI ready to go, often before you've finished your first cup of coffee.

Key features of Salesforce AI Knowledge Creation

So, once it's all set up, what does Salesforce AI Knowledge Creation actually do for your agents on a day-to-day basis? It really comes down to a few core functions.

Automated article drafting

The main event is turning conversations into articles. When an agent closes a case from a chat or email, they can click a button to have Einstein draft a knowledge article about it. This is a pretty neat way to capture solutions that might otherwise get buried in old ticket threads. It helps agents contribute to the knowledge base without having to completely step out of their main workflow.

Article revision help

Beyond creating new articles, the tool can also help clean up existing ones. An agent can highlight text in an article and ask Einstein to fix grammar, make it shorter, or just improve the phrasing. It's like having a little editing assistant built-in, which helps keep your knowledge base consistent and professional.

Grounding on Salesforce data

Like any AI, this one needs information to learn from. It’s "grounded" on the data inside your Salesforce org, which mostly means the transcripts from customer support cases. This helps make sure the articles it generates are relevant to your business.

Salesforce is also developing a feature called Unified Knowledge that's designed to pull in data from outside sources like SharePoint or Confluence. But that's usually a separate, more involved project. The basic Knowledge Creation tool is stuck with the data that's already inside Salesforce.

This brings up a pretty big question: what about all the important information your team keeps elsewhere?

This is where a more open solution can make a huge difference. While Salesforce needs a whole separate project to connect to outside knowledge, tools like eesel AI are built to connect to everything from the start. You can instantly plug into Confluence, Google Docs, Notion, and tons of other apps. eesel AI also trains on thousands of your past tickets automatically, so it develops a deep understanding of your customer issues that goes way beyond a single conversation.

Limitations of Salesforce AI Knowledge Creation

While drafting articles from cases is certainly useful, the tool has some real-world limitations you should think about before you commit.

Locked into the Salesforce ecosystem

Salesforce AI Knowledge Creation is built to work inside the Salesforce bubble, and that's about it. If your detailed product guides are in Confluence, your internal troubleshooting docs are in Google Docs, and your support team uses a different help desk, this feature won't do you much good. It keeps your information in separate silos instead of bringing it all together. You're left with two choices: move everything into Salesforce (a massive project) or pay for the more complex Unified Knowledge solution.

Reactive, not proactive

The tool is designed to react. It creates a draft article after a conversation has already happened. It can't look across thousands of tickets to find trends and tell you where your biggest knowledge gaps are. You're still depending on individual agents to notice that a new article is needed, one ticket at a time.

In contrast, a tool like eesel AI gives you reports that show you what your customers are actually asking about. It flags the common questions that don't have a good answer in your knowledge base, giving you a data-driven to-do list for new content. It can even generate draft articles from past ticket resolutions to help you fill those gaps much faster.

No way to test before you go live

Rolling out a new AI tool can feel like you're just hoping for the best. How do you know if it will be accurate? Will it work well for your specific customer questions? The Salesforce documentation doesn't mention a simulation mode that lets you test the AI's output on your old tickets before you launch it. This makes it hard to predict how well it will perform or what your return on investment will be.

This is another spot where more modern platforms give you more control. For example, eesel AI has a simulation mode that lets you test your setup on thousands of your past tickets in a safe environment. You can see exactly how the AI would have answered, forecast how many tickets it could resolve, and adjust its behavior before it ever interacts with a real customer. This lets you roll out automation slowly, starting with simple queries and expanding as you get more comfortable.

Pricing and packaging

Trying to figure out how much Salesforce Einstein features cost is famously tricky. You won't find a simple price on their main pricing page. Tools like AI Knowledge Creation are often bundled into their most expensive plans, like the Unlimited Edition, or sold as part of an "Einstein 1" add-on.

Most of the time, you'll have to get on a call with their sales team to get a custom quote. This lack of clear pricing can make it hard to budget and often means the total cost is much higher than you'd guess. You're not just paying for one feature; you're paying for a whole suite of services you might not even need.

A more flexible alternative to Salesforce AI Knowledge Creation: eesel AI

If the idea of being locked into one ecosystem with a complicated setup sounds like a headache, it might be time to look at another option. eesel AI is a self-serve AI platform built to work with the tools you already have, not make you switch everything over.

Here’s a quick comparison:

• Go live in minutes: Get set up with a single click. Just connect your help desk, point it to your knowledge sources, and you’re good to go.

• Unify your knowledge: Instantly connect to help desks like Zendesk or Freshdesk, docs in Confluence and Google Docs, and your past ticket history.

• Test with confidence: Use the simulation mode to test your setup on historical data so you know what to expect when you go live.

• Transparent pricing: You know exactly what you're paying for with clear plans and no surprise fees.

Plan	Monthly (bill monthly)	Effective /mo Annual	Bots	AI Interactions/mo	Key Unlocks
Team	$299	$239	Up to 3	Up to 1,000	Train on website/docs; Copilot for help desk; Slack; reports.
Business	$799	$639	Unlimited	Up to 3,000	Everything in Team + train on past tickets; MS Teams; AI Actions (triage/API calls); bulk simulation; EU data residency.
Custom	Contact Sales	Custom	Unlimited	Unlimited	Advanced actions; multi‑agent orchestration; custom integrations; custom data retention; advanced security / controls.

Final thoughts on Salesforce AI Knowledge Creation

So, what's the verdict? Salesforce AI Knowledge Creation can be a decent tool for teams that are deeply embedded in the Salesforce ecosystem. It provides a handy way to capture knowledge from support conversations and keep your knowledge base from getting stale.

However, its reliance on the Salesforce platform, reactive approach, complicated setup, and murky pricing make it a tough sell for many support teams. If you’re looking for a solution that’s flexible, easy to roll out, and connects to the tools you already love, you’ll probably get more out of a dedicated AI platform. The best tools today should adapt to your workflow, not the other way around.

Ready to connect your knowledge and automate support in minutes, not months? Try eesel AI for free.