Issue with Passing Parameters from "Page" to "Playbook" in Conversational Agent project
Hi everyone,I'm currently working on a conversational agent to create a seamless chat experience. My goal is t...
AI/ML
Have questions around Artificial Intelligence or Machine Learning? This is the place to ask questions about all things AI and ML, including products like AutoML, Vertex AI, AI Infrastructure, and more.
Forum Posts
Issue Passing Parameters from Flows to Playbook
Hello, I am working on passing the parameters from Flows to Playbook, I tried with all different methods and b...
Dialogflow Chatbot SSL Security Warning (CX)
Hello everyone,I'm getting this error whenever I try the demo of chatbot...it does not load and says in the co...
Dialogflow Messenger for Dialogflow Essentials not appearing
Hi Team, recently seen a post some hours ago regarding this situation, I have checked on my side and Dialogflo...
Dialogflow ES training very slow to complete
Hello! When I initiate training on my projects I am getting a very slow response. The training starts and you ...
Support for Google Sheets as a source to NotebookLM
Hope this is the right board as did not find one for genAI or Gemini or NotebookLMWe have users that want insi...
Document AI data storage
Hello,I want to process DOCX/PDF files stored in GCP storage with Document AI full OCR processor and at run ti...
Imagen3 material inpaint
Hi,Currently i'm playing with the python genai package for Imagen3 image editing.Although I can apply prompts ...
Fine Tuning Llama2
Help me guys for deploying llam2 fine tune in vertex ai . i have t4 gpu support with tesla . even wit the samp...
Others may have made recent changes to the example.
Keep getting this error “ Others may have made recent changes to the example. Please refresh your CX console t...
Custom metadata labels for embedding models
Vertex generative models provide custom metadata labels (for requests) but embedding models don't. Does anyone...
Resource Exhaustion Errors (429) when using grounding via google_search tool
I am on a paid plan and have been using the grounding feature with gemini-2.0-flash (config["tools"] = [{"goog...
Consistent 429 even if your metrics says I am well under free tier quota
I've integrated gemini api into a project of mine and sending prompts with images.I am on the free tier, and t...
Journey / Chirp HD TTS is very slow to respond... ~ 4,000ms. Not a production ready API from google.
So, for a single sentence I'm getting latencies of nearly 3,000ms... Is this normal? I've noticed Google quali...
Start an executor from a notebook that has dependencies on other files
We have a notebook A.ipynb which imports a module B.py.Running A.ipynb from the Vertex AI Workbench works fine...
Dialogflow Es messager is not working
I am using Dialogflow Messenger on my site, but now it is not working. I also tried it directly on the site, b...
Imagen 3.0 Model Failure Pricing Implications
On the odd occasion imagen 3.0 fails to generate an image when using the API due to resource exhaustion or gen...
Dialogflow CX 32601 Error
I'm feeling a bit desperate with my Dialogflow agent. I'm connecting to it through Twilio. I make calls, and i...
No option to ask to increase Imagen3 quota increase in Vertex AI
Hi google!I still don't see an option to request to increase imagen 3 limitations and this is very important t...
Gemini 1.5 Flash: API Results Worse than AI Studio?
Hey y'all. Pretty new to Gemini, so pardon ahead of time if I'm just being an idiot. I do that sometimes.Probl...
AutoML Image classifier taking days for training
I started to train a AutoML image classifier on Vertex AI. The data set is just about 36 images of dogs and fl...
Can you reference input parameters in a playbook example (Conversational Agents)
In a Conversational Agents playbook that is called from a flow with input parameters including age and locatio...
Vertex AI Workbench
I"m trying to provision Vertex AI workbench using the already existing service account using terraforming scri...
Google Vision OCR randomly not detecting negative signs
Hi All,I am trying to extract values from a financial document using Google Vision OCR and for an unknown reas...
Dialog flow message is not working in demo
I am using Dialog flow Messenger on my site, but now it is not working. I also tried it directly on the site, ...
Error de despliegue - Librerias SSL
Estamos experimentando un error con los bots de Dialogflow al momento del despliegue. Según estuvimos investig...
Gemini 1.5 Pro Training on Vertex - Dataset Formatting from Google AI Studio ready data.
i have a google ai studio compatible dataset, I want to train gemini 1.5 pro on vertex on the same dataset, wh...
Speech-to-Text. Unexpected Transcribing Numbers as Digits
Google "Speech to text" solution started transcribe numbers as digits instead of words. For example, return "1...
Regarding accurate detection of Cantonese Language CCAI voice enabled Dialogflow ES bot
Problem encountered: The Cantonese language is either not getting detected or is being detected incorrectly in...
how i can get technical support?
I am trying for more than two days to open a technical ticket and is not possible.....What I should try furthe...
Articles from Product Experts and Googlers
Empowering conversational interfaces with BigQuery analytics
In our previous blog posts (1, 2, 3 and 4), we explored how Agent Builder can intelligently answer questions from BigQuery and unstructured data stores via chat or search bot. However, things get trickier when we need to address analytical queries that involve aggregations, comparisons, and more complex conditions.
From "smol" to scaled: Deploying Hugging Face’s agent on Vertex AI
Building and scaling your own AI agent can seem complex, but it doesn't have to be! HuggingFace’s smolagent library provides a lightweight framework to build powerful agents.
Working with BigQuery JSON & nested data
JSON (JavaScript Object Notation) is a lightweight, text-based data format designed for human readability and easy machine parsing. A big advantage of using JSON data in BigQuery is that BigQuery treats JSON as semi-structured data, meaning you don't need a fixed schema upfront.
