- 1095 Views
- 5 replies
- 4 kudos
Resolved! How do you test whether a retriever stops too early?
Hello everyone,Databricks’ Adaptive Instructed-Retriever announcement describes a model that learns when to stop searching. How would you test whether it stops after finding a general policy but before finding the exception that changes the answer?I...
- 1095 Views
- 5 replies
- 4 kudos
- 4 kudos
Hi @ivanvyd, sorry for the slow reply. Short honest answer: I have not run paired questions against the Adaptive Instructed-Retriever itself, so what follows is how I would set the test up, not results. Paired design, the way I would build it: take o...
- 4 kudos
- 1064 Views
- 6 replies
- 6 kudos
Resolved! Genie Agent SDK visualization
The documentation says you can send in an optional enable_visualization boolean when starting a conversation or sending a message to the agent:start_conversation_and_wait(space_id: str, content: str [, enable_visualization: Optional[bool], timeout: d...
- 1064 Views
- 6 replies
- 6 kudos
- 6 kudos
Hello @haaland, I took a look at both internal and external documentation and here is what I found. Glad it's working, and @Gecofer and @ivanvyd had the right diagnosis. enable_visualization was never removed. It shipped in databricks-sdk v0.120.0 (J...
- 6 kudos
- 770 Views
- 3 replies
- 1 kudos
Resolved! Genie Agents citation (bug)
Agent reponses are showing this kind of plain text at the end of its text reponses: :citation[genie_sql.01f1b7e5-18d5-196a-90f2-2601ef936192]I think it is meant to be a clickable link. Will this be fixed soon?
- 770 Views
- 3 replies
- 1 kudos
- 1 kudos
Hi @haaland - I believe the fix has now been rolled out, and I have tested my sandbox workspace where citations are working correctly. Please check and let me know if you still face issues. Happy to follow up internally.
- 1 kudos
- 4496 Views
- 2 replies
- 1 kudos
Resolved! How to deploy an Agent
Hi everyone, I’m currently developing several Databricks Agents using a custom Claude skill to integrate Databricks Asset Bundles (DABs), Mosaic AI frameworks, and LangGraph. My current workflow is the following: I develop the agents locally.For depl...
- 4496 Views
- 2 replies
- 1 kudos
- 1 kudos
Hey, your approach is a common pattern. But there has been a recent shift where we now recommend creating the agents as databricks apps instead. This allows you to not do pipeline execution steps. Here are a couple of articles you may find helpful. h...
- 1 kudos
- 7105 Views
- 11 replies
- 0 kudos
Can't Access Agents
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 ...
- 7105 Views
- 11 replies
- 0 kudos
- 3517 Views
- 1 replies
- 1 kudos
Resolved! Model Serving Endpoints Scaling From Zero Forever
Hello,When deploying agents via a Serving Endpoint with "Scale To Zero" enabled, we are encountering that at seemingly random times, when we hit the endpoint to wake it up, it will begin to "Scale From Zero", but will never become fully available. Th...
- 3517 Views
- 1 replies
- 1 kudos
- 1 kudos
Hi @jAAmes_bentley, There are multiple possible root causes, so let me walk through them so you can diagnose which applies to your situation. WHAT HAPPENS DURING SCALE-FROM-ZERO When an endpoint scales from zero, Databricks must: 1. Acquire comp...
- 1 kudos
- 837 Views
- 1 replies
- 1 kudos
Ai query parallel calls
I’m trying to optimize ai_query calls on a table and wanted to get some ideas.So far, I’ve tried repartitioning the DataFrame before running spark.sql(ai_query), but I didn’t see any meaningful performance gains. I also experimented with running mult...
- 837 Views
- 1 replies
- 1 kudos
- 1 kudos
When you are using ai_query(), there are two main aspects to performance: Model serving endpointSQL warehouse / Compute cluster Very likely, the performance is throttled by the model-serving endpoint's concurrency limit. Reference: https://docs.d...
- 1 kudos
- 1637 Views
- 3 replies
- 4 kudos
Multi‑Agent Supervisor: url_citation (source links) not shown in Playground — why and how to enable?
I’m seeing a difference in citation behavior between a single Knowledge Assistant (KA) agent and the Multi‑Agent Supervisor setup.What I testedIn Agent Bricks → Knowledge Assistant, I created an agent that returns citations with links (e.g., url_cita...
- 1637 Views
- 3 replies
- 4 kudos
- 3256 Views
- 3 replies
- 5 kudos
Resolved! Information Extraction Agent - Update Agent failed
Hi everyone,I’m running into a persistent issue when trying to save an Information Extraction Agent in Databricks.Every time I click Save and update, I get: Update Agent failed. Failed to save AgentThis happens regardless of what I change — schema, g...
- 3256 Views
- 3 replies
- 5 kudos
- 5 kudos
It sounds like the update for your Information Extraction Agent didn’t go through successfully. A few things you could try:Check logs – Look for any error messages during the update process; they often give clues about what went wrong.Verify dependen...
- 5 kudos
- 3332 Views
- 3 replies
- 0 kudos
Resolved! Testing out Agentic Capabilities
So I am creating a POV on Databricks' agentic capabilities and wanted to showcase its abilities through a simple change pipeline.A user asks for changes in a specific table in a schema -> based on metadata info from our lake table info is received ->...
- 3332 Views
- 3 replies
- 0 kudos
- 0 kudos
Yes, it’s viable — Databricks’ in-house agents can handle that workflow if you define each stage as callable Python tools. The key is robust function design and proper metadata access. However, you might find integrating Unity Catalog and Git integra...
- 0 kudos
- 3705 Views
- 4 replies
- 2 kudos
AI Agents - calling custom code and databricks jobs
Hi EveryoneI am building AI Agents; my requirement is to call custom tool logic (which was not possible using a unity catalog function) and databricks jobs. I could not find much of documentation on these scenarios. If someone could share any referen...
- 3705 Views
- 4 replies
- 2 kudos
- 2 kudos
Hi @GiriSreerangam , can you share what you are trying to do with the custom tool? You might be able to implement with a custom MCP server. Here are some other resources that may help: https://github.com/JustTryAI/databricks-mcp-serverhttps://github...
- 2 kudos
- 3231 Views
- 5 replies
- 5 kudos
Resolved! Databricks Free Edition - AI Agent Export Limitation
Hi, I am following the Databricks AI Agent guide to create a custom agent using my functions. However, in the Free Edition, I do not see the option to export the agent from the Playground. Is this feature unavailable in Free Edition, or is there a wo...
- 3231 Views
- 5 replies
- 5 kudos
- 5 kudos
Ok @atharvaghodekar , so there's no export button (neither on premium workspace as @WiliamRosa stated above).I think they refactor UI and now the same functionality is under Tools -> Create agent notebook. So, in tutorial export button is reposonsib...
- 5 kudos
- 5812 Views
- 4 replies
- 3 kudos
LangChain Compatibility in Databricks – Painfully Inconsistent Experience
Hey all,Is anyone else experiencing major friction when using LangChain within Databricks?Even after carefully pinning versions like:%pip install \ loguru \ mlflow==2.10.1 \ langchain==0.1.5 \ langchain-google-vertexai \ databricks-vectorse...
- 5812 Views
- 4 replies
- 3 kudos
- 3 kudos
I'll definitely look into installing libraries at the cluster level via init scripts makes a lot of sense for stability and reproducibility. I hadn't considered that approach, so I really appreciate you highlighting that and providing the relevant Da...
- 3 kudos