Sample Bots Introduction

The ACE Agent Quick Start package comes with a number of sample bots which can be found in the ./samples/ directory. The sample bots are built for different use cases and industries and showcases various features of ACE Agent. You can build your own bot on top of the sample bots or build from scratch by following the tutorials section.

Sample bots demonstrate specific features of the ACE Agent pipeline. We have highlighted a few features in the following table. Perform the steps in the individual bot sections for getting details on the full list of features supported. Not all bots support all features. For bot customizations, refer to the Tutorials section based on your use case.

List of Sample Bots

Sample Bot	Use Case	Features
Chitchat Bot	Chitchat bot can have informal and smalltalk conversations with the user	Using LLM for conversation with Colang Supports Event Architecture and Chat Engine Server Architecture
LLM Bot	Showcase how LLM can be hooked in the ACE Agent Speech pipeline	Integrating LLM / Agents using Colang Action ASR 2 pass EOU for speech latency reduction Always-on any word Barge-In Handling conversation history in Colang Actions Supports Event Architecture
RAG Bot	Sample bot showcasing integration with NVIDIA Generative AI RAG Examples	Integrating RAG solutions using the Plugin Server ASR 2 pass EOU for speech latency reduction Always-on any word Barge-In Handling conversation history in plugins Supports Event Architecture
Stock Bot	Sample bot for queries related to stock market and fetching live stock prices	Plugin Server for external APIs integration such as Yahoo finance Handling profane questions Answering contextual questions Supports Event Architecture and Chat Engine Server Architecture
Gaming Non-Playing Character (NPC) Bot	LLM Driven Gaming Non-Playing Characters	Integrating Small language model (SLM) with the bot Custom SLM prompts for each NPC character ASR 2 pass EOU for speech latency reduction Always-on any word Barge-In Supports Event Architecture
DuckDuckGo LangChain Bot	LangChain agent that uses conversation history and the DuckDuckGo tool to answer questions	Integrating a LangChain agent without using Colang in Speech pipeline Handling conversation history in the Agent Installing custom dependencies in the Plugin server Supports Plugin Server Architecture
Spanish Weather Bot	Spanish weather bot utilizing Riva Translation models or LLM models	Using Riva NMT for translations and LLM with English prompt Multilingual LLM prompt to Spanish input and provide Spanish bot response Supports Event Architecture and Chat Engine Server Architecture
Food Ordering Bot	Virtual assistant for Food Ordering domain	Showcase legacy features, such as, intent slots models and slot management along with LLM conversation Support Event Architecture and Chat Engine Server Architecture
LangGraph Bot	Sample bot that showcases the Plan & Execute LangGraph pipeline integration that uses LLM and Tavily search to answer complex questions	Integrating LangGraph agent /runnable with ACE pipeline Supports Event Architecture and Plugin Server Architecture

In the sections below, we will showcase deployment using the Docker Environment only, but the same should be possible via the Kubernetes Environment and for text bots using the Python Environment. The ACE Agent Quick Start package comes with a number of UCS applications for sample bots to support kubernetes deployment which can be found in the ./deploy/ucs_apps/ directory.