01-16-2026 12:30 AM
Hi there, I'm very new to Databricks and I wanted to try out Agents. At first, I thought I couldn't access Agents because I was on a Trial account, so I added my CC to the subscription. Unfortunately, I still can't access the feature, not to mention my account already owes $30 without even getting to run anything 🙁.
I couldn't find any direct Databricks support, so trying my luck here. Can anyone point me in the right direction? I'm using us-east-1 region on AWS with Databricks-managed backend.
01-16-2026 01:00 AM
Hi @timbaron ,
This feature is still in Beta. First of all check if you have enabled it in Priviews (as they suggest in docs):
Use Agent Bricks: Information Extraction - Azure Databricks | Microsoft Learn
Note, that access to some Beta features could be gated. Then you need to contact with databricks team and ask them directly to enable this feature for you.
Click send feedback and you will find a textbox where you can describe your issue. There you can also find Help Center section - it could be useful to find another means of contact:
01-16-2026 01:05 AM
Also, check if you're meeting all requirements for Agent Bricks:
Agent Bricks - Azure Databricks | Microsoft Learn
01-18-2026 11:40 AM
Hi Szymon, thanks for a prompt reply!
I followed your suggestions (AWS version) but sadly it seems my account is not on the allowlist for Mosaic Beta.
I messaged Databricks support via email, but it seems I also don't have access to the Help Center either.
If you or anyone else has any suggestions on how to make this work, I'm all ears. 🙏
2 weeks ago
Hi @timbaron,
There are a few things to check here that will help you get Agents working on your workspace.
WORKSPACE SUBSCRIPTION TIER
Mosaic AI Agent Framework features (including Agent Builder, AI Playground, and model serving for agents) require a Premium or Enterprise tier workspace. If your workspace is on the Standard tier, these features will not appear in the UI. You can check your current tier by going to your Account Console and looking under the subscription or billing section. If you are on a Standard plan, you would need to upgrade to Premium.
ENABLE PREVIEWS
Some Agent-related features are delivered as Previews (Beta or Public Preview). To enable them:
1. Click your username in the top bar of your Databricks workspace.
2. Select "Previews" from the menu.
3. Look for any AI or Agent related toggles and make sure they are turned on.
If you are an account admin, you can also check account-level previews:
1. Log in to the Account Console.
2. Click "Previews" in the sidebar.
3. Enable any relevant AI features.
Note that some features may be in Private Preview, meaning they are invite-only and require contacting your Databricks representative. If a toggle does not appear at all, that feature may be gated for your account.
REGION SUPPORT
You mentioned you are on us-east-1 (AWS), which is good. That region supports Mosaic AI Agent Framework, Agent Evaluation, Foundation Model APIs (pay-per-token and provisioned throughput), and model serving. So region is not the issue here.
UNITY CATALOG
Agents need to be registered in Unity Catalog before deployment. Make sure Unity Catalog is enabled on your workspace. You can confirm this by checking if you see a "Catalog" icon in the left sidebar of your workspace.
THE $30 IN CHARGES
When you add a credit card and your workspace becomes active, certain background resources can incur charges even before you run anything explicitly. This can include things like serverless compute warming up, or if any clusters were left running. To investigate:
1. Go to your Account Console.
2. Check the "Usage" section to see what DBUs were consumed and by which resources.
3. Make sure you do not have any clusters running that you are not using. You can check under Compute in your workspace.
GETTING SUPPORT
Since you mentioned you could not access the Help Center, here is a direct path:
1. In your Databricks workspace, click the "?" icon in the top bar.
2. Select "Support" or "Send Feedback."
3. Describe your issue there, specifically that you cannot access Agent Framework features after upgrading from trial.
You can also email Databricks support directly at help@databricks.com.
DOCUMENTATION REFERENCES
Here are the key docs for getting started with agents on AWS:
https://docs.databricks.com/aws/en/generative-ai/agent-framework/build-genai-apps.html
https://docs.databricks.com/aws/en/generative-ai/agent-framework/create-agent.html
https://docs.databricks.com/aws/en/generative-ai/agent-framework/deploy-agent.html
https://docs.databricks.com/aws/en/admin/workspace-settings/manage-previews.html
The most likely culprit is either the workspace tier or a preview toggle that needs to be enabled. Start with checking those two items and you should be on your way.
* This reply used an agent system I built to research and draft this response based on the wide set of documentation I have available and previous memory. I personally review the draft for any obvious issues and for monitoring system reliability and update it when I detect any drift, but there is still a small chance that something is inaccurate, especially if you are experimenting with brand new features.
If this answer resolves your question, could you mark it as "Accept as Solution"? That helps other users quickly find the correct fix.
2 weeks ago
Same issue here. Azure Databricks Premium PAYG, East US region, all prerequisites met
(Unity Catalog, serverless, system.ai with 56 models, budget policy, all previews
enabled). Knowledge Assistant and Supervisor Agent both greyed out with 'contact
sales' error. Workspace URL: adb-7405613569341731. Submitted feedback 2026-03-11, no
response yet.
2 weeks ago
Hi Badr,
Since you are on Premium PAYG in East US with Unity Catalog, serverless, system.ai, and budget policy all set up, your workspace should qualify.
That said, there are a couple of things worth double-checking:
If all of the above checks out and you are still seeing the "contact sales" error with greyed-out options, this may be a provisioning issue on the backend. Since you already submitted feedback on 2026-03-11, I would also recommend opening a support ticket referencing workspace adb-7405613569341731 so the team can check whether the feature flag is properly enabled for your workspace.
Reference:
2 weeks ago
Hi Steve, thanks for the response.
To confirm both points:
1. Preview enablement — confirmed ✅
Mosaic AI Agent Bricks Preview (Beta) is explicitly toggled on in Admin Console →
Workspace Settings → Previews. I double-checked — it's enabled.
2. Serverless budget policy — this is where it gets unclear
I created budget policies in two places, but neither UI offered a way to specify a
nonzero budget amount:
- Workspace level: Settings → Compute → Policies → Serverless usage policies → Manage
→ Create + tag. The form doesn't have a field for setting a budget amount.
- Account Console level: Usage → Budgets → Create Budget. Same — I created one, but
the UI didn't ask for a dollar amount or nonzero constraint.
The docs say "Access to a serverless budget policy with a nonzero budget", but the
actual UI for creating these policies doesn't seem to surface a field where you set a
numeric budget value. So I'm not sure how to satisfy this requirement — or how to
verify whether my current policies count as "nonzero."
Could you clarify:
- Where exactly do you set the budget amount for a serverless budget policy?
- Is there a way to verify the current budget value via the UI or CLI?
If all of the above checks out and this is indeed a backend provisioning issue, I've
already submitted feedback on 2026-03-11 referencing workspace adb-7405613569341731.
I'll open a formal support ticket as you suggest.
Thanks for looking into this.
2 weeks ago
Hi Badr,
Thanks for the detailed follow-up -- this is a genuinely confusing area of the docs, so you are not alone in being tripped up by it.
The short answer: serverless budget policies (Settings > Compute > Serverless budget policies > Manage > Create) are tagging mechanisms for cost attribution, not spending caps. They do not have a "budget amount" field -- you just give them a name and tags. That is expected behavior in the UI you saw.
The account-level budgets (Account Console > Usage > Budgets > Add Budget) are separate -- those do have a dollar amount field and are used for monitoring spend. But they are also alert-only and do not block access.
The docs requirement that says "Access to a serverless budget policy with a nonzero budget" is unfortunately worded in a way that conflates these two concepts. In practice, the requirement is satisfied by simply having a serverless budget policy created and having your user assigned to it (with the "User" or "Manager" role on the policy's Permissions tab). There is no numeric budget value you need to set within the policy itself.
So to confirm your setup:
If all of that checks out and you still cannot access Agents, the support ticket referencing workspace adb-7405613569341731 is the right move -- it may be a backend provisioning issue that the product team needs to investigate.
Sources:
2 weeks ago
Thanks for the clarification on budget policies vs account-level budgets — that clears
up a lot of confusion. The docs wording really does conflate the two.
To confirm, I've now verified everything:
- Serverless budget policy: Created at workspace level (Settings → Compute →
Serverless budget policies). My user is assigned with the "User" role on the
Permissions tab. ✅
- Serverless compute: Enabled ✅
- Unity Catalog: Enabled ✅
- Workspace region: East US ✅
- Mosaic AI Agent Bricks Preview: Enabled in Previews ✅
Still greyed out — Knowledge Assistant and Supervisor Agent both show "contact sales."
Regarding the support ticket: I'm on a self-service Premium PAYG account, so I don't
have access to help.databricks.com to open a formal support ticket. The only option I
had was submitting feedback through the workspace UI, which I did on 2026-03-11
referencing workspace adb-7405613569341731.
Is there another way to escalate this for PAYG accounts without access to the support
portal? Or is this something you could flag internally to the product team?
Thanks again for your help.
2 weeks ago
Do you have Microsoft support for your Azure presence? If so, you can submit a ticket to them for Azure Databricks. They are first-line support for us on Azure since Azure Databricks is a first-party service.
2 weeks ago
no support whatsoever