Activity
Mon
Wed
Fri
Sun
Oct
Nov
Dec
Jan
Feb
Mar
Apr
May
Jun
Jul
Aug
Sep
What is this?
Less
More

Memberships

AI Developer Accelerator

9.9k members • Free

4 contributions to AI Developer Accelerator
Google ADK "supervisor" architecture
Hi everyone, I’m currently testing Google’s ADK and building my first multi-agent system (a customer service agent). My idea is to have a root agent that communicates directly with the customer, while the sub-agents only interact with or assist the root agent (not the customer directly). The problem I’m facing is that all agents are responding to the customer, which breaks some intended features. For example: - The root agent can open a ticket using a tool when it cannot solve the customer’s problem. - If the customer’s prompt is about an order and the root agent delegates it to the “orders” sub-agent, the sub-agent should send its answer back to the root agent. - However, what happens now is that the sub-agent replies directly to the customer saying it can’t do anything, instead of sending this information to the root agent, which should then decide whether to answer or propose opening a ticket. Has anyone faced this issue before or knows how to make only the root agent respond to the customer?
1 like • Aug 21
@Darnel C I’m using Google’s RAG Engine for this. I upload the contexts as PDFs from Google Drive, keeping the default settings for chunking, precision, etc. (so far they give good results, but I plan to experiment with other variations). For querying, I built a custom tool similar to the one in this video, but with a twist: the client doesn’t have direct access to the corpora. Instead, I define them in code so that each query tool only has access to the corpus it needs. This way I keep some “opacity” and naturalness, giving the client the impression that the agent just knows all that information because it’s an expert 😉 Here’s the video I based my query tool on. If you’re curious, I can share some code with you later, just let me know 😀 https://youtu.be/TvW4A0a75mw
1 like • Aug 21
@Darnel C By the way, I’m curious about your side too, what are you exploring with Document AI?
AI with agents and tools
Hello All as I stepped into AI space few months ago, am trying to work on agent which can do code analysis any application started with Java. So far I have agent for pulling data from repo Store/ clone to some location Another agent to analyze the codebase Am trying to stick them together Any pointers ? Am using google adk
0 likes • Aug 14
As far as I know, there are two ways to “connect” two agents: one being a sub-agent of the other, or using a central agent and then building an agent-as-a-tool that the main agent can use. I should probably take the second option: one agent to which you pass the link (or the option you chose) of the repo, and then use AgentTools to do the tasks you are describing. I also always think about whether it’s really necessary to use an agent for some tasks or if it could be a simple tool (sometimes I go too fast to build an agent, and sometimes it’s not necessary!). What do you think? did it help?
0 likes • Aug 14
https://google.github.io/adk-docs/tools/function-tools/#key-aspects-of-this-example
Building a Multi-Agent Product Recommender with Google ADK & Vertex AI
Hello, I'm building a multi-agent system with Google ADK. I have seen what Zalando built (https://www.cxnetwork.com/replay-video/content/zalando-ai-assistant) and I think it is awesome. Now I'm wondering what tools they used to build an agent like that. Can we, with Google ADK, Vertex AI, and Google's toolset, build something like this? What do you think? what would be the way to do it? I have been investigating, and I should probably build a RAG agent with access to a high-quality document or set of documents with a very specific guide of recommendations, history of products, etc. (all valuable information to recommend products). I would also create a product database using a Google tool for that, accessible by the agent so it can search products by all parameters and view images of the products. (Before choosing products, it should verify the price and stock with the API I'm using in my e-shop). What do you think? What tools would you use? Is Google ready for a chatbot like Zalando's?
Hi! I'm Albert
I’m really excited to be here and look forward to learning from all of you, sharing knowledge, and helping each others. See you here!
4
0
1-4 of 4
Albert Esponey
2
7points to level up
@albert-esponey-4778
I'm a full stack web developer with a few years of experience, now exploring the potential of AI

Active 9d ago
Joined Aug 7, 2025
Powered by