- 41862 Views
- 24 replies
- 51 kudos
Databricks Announces the Industry’s First Generative AI Engineer Learning Pathway and Certification
Today, we are announcing the industry's first Generative AI Engineer learning pathway and certification to help ensure that data and AI practitioners have the resources to be successful with generative AI. At Databricks, we recognize that generative ...
- 41862 Views
- 24 replies
- 51 kudos
- 51 kudos
Great initiative by Databricks! This is a valuable opportunity to build practical GenAI skills and stay ahead in the rapidly evolving AI space.
- 51 kudos
- 1511 Views
- 1 replies
- 0 kudos
Resolved! AI playground - Unable to access LLM's
I was having playground access and now I bumped up to enterprise access The endpoint GPT-5.4 has encountered an error and responded with the following: InternalError{"error_code":"PERMISSION_DENIED","message":"PERMISSION_DENIED: The endpoint is tem...
- 1511 Views
- 1 replies
- 0 kudos
- 0 kudos
Hi @nagamaddikunta, Having checked this internally, this error message doesn’t come from the QPM/TPM values you see on the Serving endpoint. It means that, for your workspace, Databricks has set the system-level quota for that model to 0, so the GPT-...
- 0 kudos
- 2712 Views
- 3 replies
- 2 kudos
Resolved! Knowledge cutoff for Genie
Hi, I was trying to make a dashboard using AI Genie - it worked well for the basics, but was unable to perform some of the modifications that were obviously (via UI) doable. Per Genie's response below - it only knows Databricks documentation up to Ap...
- 2712 Views
- 3 replies
- 2 kudos
- 2 kudos
Hi, Just wanted to add to what others have said and make sure its clear, I think you're trying to use a Genie Space to create your dashboard. But Genie code is probably a better tool here. If you have visuals created in a Genie space that you would l...
- 2 kudos
- 992 Views
- 2 replies
- 2 kudos
Resolved! Request for blog publishing access
Hi,I would like to request access to publish blogs on the Databricks Community. I am a seasoned SME in Generative AI, with extensive hands-on experience in this domain. I have actively worked on various GenAI use cases and regularly write about emerg...
- 992 Views
- 2 replies
- 2 kudos
- 2 kudos
Hi @mansi26 ,Pleasure to meet you. You can publish your articles without asking for consent in community articles section:https://community.databricks.com/t5/community-articles/bd-p/Knowledge-Sharing-Hub
- 2 kudos
- 6481 Views
- 1 replies
- 0 kudos
Resolved! Error: "Invalid model name" in Databricks AI Gateway when setting up Vertex AI endpoint
Hi everyone,I'm trying to set up a new serving endpoint in Databricks using Google Cloud Vertex AI as the provider. I want to route it to claude-opus-4-6.However, as soon as I try to create it, I get the following UI error (see screenshot):"Invalid m...
- 6481 Views
- 1 replies
- 0 kudos
- 0 kudos
Hi @martkev -- good news: your model name is correct. The Vertex AI model ID for Claude Opus 4.6 is indeed claude-opus-4-6 (per Anthropic's official documentation). The issue is on the Databricks side -- the UI enforces an internal allowlist of recog...
- 0 kudos
- 990 Views
- 1 replies
- 0 kudos
Petition to update your documentation about inferencing LLMs and standardize LLM response format
TL;DRInconsistent & Undocumented Response Format: Contrary to Databricks documentation, enabling thinking.type for Gemini 2.5 Flash changes the ChatCompletion content field from a standard string to a list of dictionaries, breaking expected behavior....
- 990 Views
- 1 replies
- 0 kudos
- 0 kudos
Hey @trickywhitecat , You're right that when thinking.type is enabled for Gemini 2.5 Flash on a Databricks serving endpoint, the content field comes back as a list of dictionaries instead of a plain string. That breaks the expected OpenAI ChatComplet...
- 0 kudos
- 3805 Views
- 1 replies
- 2 kudos
Resolved! ai_query and cached tokens
Is ai_query actually able to use OpenAI's cached tokens? I was not unable to prove it. The response object from ai_query does not contain the raw response, and when I re-run an identical request via OpenAI SDK (identical model, settings etc.) and exa...
- 3805 Views
- 1 replies
- 2 kudos
- 2 kudos
Great question -- this is a nuanced topic because there are two layers involved: Databricks' proxy layer and OpenAI's caching mechanism. Short answer: No, ai_query does not currently support OpenAI's prompt caching. 1. ai_query doesn't expose token u...
- 2 kudos
- 2859 Views
- 3 replies
- 0 kudos
Tracing through model serving endpoint
i have deployed a code running on langgraph through model serving endpoint. I want to trace the logs using ml flow and i want to trace logs in the experiment whenever a user hits the serving endpoint. I have defined both of them in my codemlflow.set...
- 2859 Views
- 3 replies
- 0 kudos
- 0 kudos
Hi @srijan1881, The behavior you are seeing is expected when using a manually created model serving endpoint rather than one deployed through the Databricks Agent Framework. Here is a breakdown of why traces are not appearing and how to resolve it. U...
- 0 kudos
- 3102 Views
- 4 replies
- 0 kudos
Resolved! Genie integration in Dashboards fails with "GenericChatCompletionServiceException" (Free Edition)
I am experiencing a persistent error when using Genie Space integration within a Databricks Dashboard (v3/Lakeview). While Genie works perfectly as a standalone Space, it consistently fails when invoked via the Dashboard sidebar.Environment:Edition: ...
- 3102 Views
- 4 replies
- 0 kudos
- 0 kudos
Yes 100%, any user would need to have select permissions on the table that the Genie Space is created on top of.
- 0 kudos
- 16592 Views
- 4 replies
- 6 kudos
Managed MCP Server for Visual Studio Code and GitHub Copilot?
Hi!I am starting to explore the new managed Model Context Protocol (MCP) server with GitHub Copilot. I have successfully configured it to use the DBSQL MCP Server that you currently find in AI/ML -> Agents -> MCP Servers. As also shown in this post i...
- 16592 Views
- 4 replies
- 6 kudos
- 6 kudos
Hi excavator-matt, Thanks for the follow-up and glad to hear you got Option C (PAT-based) working with Copilot and VSCode, and that you have moved to Claude Code with the official skills. Regarding the issues with Options A and B: Option A (OAuth U2M...
- 6 kudos
- 1222 Views
- 1 replies
- 0 kudos
Ethical Data Governance
Title: Why Responsible AI Needs to Be a First‑Class Engineering Practice (Not an AfterthoughAI teams are moving faster than ever — but the industry is learning that speed without governance creates real downstream risk. Most “Responsible AI” failures...
- 1222 Views
- 1 replies
- 0 kudos
- 0 kudos
Appreciate anyone who reads through this. I’m curious how teams are implementing governance controls in Databricks today — things like automated validation, model documentation, or lineage tracking through Unity Catalog. If you’ve built guardrails th...
- 0 kudos
- 4521 Views
- 1 replies
- 1 kudos
Resolved! Claude Code User Usage Tracking
I am using Databricks Model Serving as a proxy to connect Claude Code. I established the connection through Integrate coding agents (Connect coding agents to Databricks) by generating the environment configuration: { "env": { "ANTHROPIC_MODEL": "dat...
- 4521 Views
- 1 replies
- 1 kudos
- 1 kudos
Hi @JoaoPigozzo, After some investigation, I found that when you use Databricks Model Serving as a proxy for Claude Code, what you see in system.billing.usage is expected. That table is is designed for cost attribution by SKU / endpoint, not per‑user...
- 1 kudos
- 2477 Views
- 2 replies
- 0 kudos
Resolved! DatabricksVectorSearch seems to crash when served
Hey y'all !So I'm experimenting with the Databricks' DatanircksVectorSearch class in Python to serve as a tool that can be used by an agent. When I run it on a notebook, I get the following error:"[NOTICE] Using a notebook authentication token. Recom...
- 2477 Views
- 2 replies
- 0 kudos
- 0 kudos
@Louis_Frolio Are there audit logs or visibility into which service principal is being used and what access it has?Another question: With automatic authentication passthrough, do we still need to set DATABRICKS_HOST and DATABRICKS_TOKEN as environme...
- 0 kudos
- 2296 Views
- 1 replies
- 1 kudos
Resolved! Databricks Apps Streaming issue
I have .NextJS Databricks Apps with streaming enabled by setting "stream:true" in the JSON body which tells serve to return a streaming response (SSE format). Now this works just fine when I run the app locally via npm run dev but once I deploy the ...
- 2296 Views
- 1 replies
- 1 kudos
- 1 kudos
What resolved this issue is to set the following headers listed below. Content-Type: "text/event-stream"Connection: "keep-alive"Transfer-Encoding: "chunked"These headers help prevent proxy interference:Connection: "keep-alive" explicitly tells proxi...
- 1 kudos
- 1730 Views
- 2 replies
- 1 kudos
Resolved! Feedback not showing up in Genie from Copilot Studio Genie Agent
Hi,We've created an Agent using Copilot Studio for Genie and integrated with Teams Channel.The feedback there is working and we can see the reactions in the Copilot Studio Analytics.But the feedback is not going to the actual genie space, neither the...
- 1730 Views
- 2 replies
- 1 kudos
- 1 kudos
Hi @souravg, @Ale_Armillotta is right. At the moment, Genie only records feedback (thumbs up/down, "Fix it", comments) when it’s given directly in the Genie UI. The public Genie Conversation APIs that Copilot Studio/Teams use don’t expose any endpoin...
- 1 kudos
- 1994 Views
- 1 replies
- 1 kudos
Resolved! How to get MLflow OpenAI autolog traces from PySpark mapInPandas workers (and some pitfalls)
ContextI'm running an LLM pipeline on Databricks that distributes OpenAI API calls across Spark workers via mapInPandas. Getting mlflow.openai.autolog() to work on workers required solving three undocumented issues. Sharing here since I couldn't find...
- 1994 Views
- 1 replies
- 1 kudos
- 1 kudos
Greetings @Jayachithra , I did some digging and came up with some helpful tips/hints to help you along. On Issue 1 (explicit MLflow context): expected behavior once you realize that mapInPandas spawns isolated Python worker processes, not threads. ...
- 1 kudos
