AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: inquire about a tent sold at the shop
Goal: the potential buyer is looking for a tent. He asks the chatbot a question. The chatbot informs him and recommends a tent in stock in the shop (RAG in the Cosmos DB database):
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: the github
Here is the Github project to fork:
The image shows a GitHub page from the "contoso-chat" repository belonging to "Azure-Samples". Here are some details:
Page Header:
The deposit is public.
Options: Watch (25), Fork (2.8k), and Star (457).
Tabs: Code, Issues, Pull requests, Actions, Projects, Security, Insights.
Main content:
Main branch: hand.
Green "Code" button to clone or download the repository.
A list of folders and files with descriptions of the latest commits and update dates, for example:
.devcontainer: "Add Tracing (#175)" (updated last month)
src: "Add Tracing (#175)" (updated last month)
.gitignore: "Docs/oct updates v2 (#193)" (updated last week)
"About" section:
Repository Description: "This sample has the full End2End process of creating RAG application with Promptify and AI Studio..."
Related Tags: azure-cosmosdb, copilot tutorial, llmops, and more.
All in all, it's a well-organized and documented repository that's essential for collaboration and development of open source projects.
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: fork creation
The image shows a user interface to create a new fork of a repository on a source code management platform: GitHub. Here are the visible elements:
An "Owner" field with the value "azurechatiaicopilot".
A "Repository name" field with the value "contoso-chat" and a green checkmark indicating that the name is available.
An optional description section with the text: "This sample has the full End2End process of creating RAG application with Prompty and AI Studio. It includes C".
A checkbox to copy only the "main" branch.
A message that the user is creating a fork in their personal account.
A green "Create fork" button at the bottom right.
This image illustrates the process of creating a fork, allowing users to experiment on a repository without affecting the original project.
On the "Create a new fork" page, scroll down and uncheck the "Copy only the main branch" option. If you forget to uncheck this option, you will have to delete your fork and try again.
Click on the "Create Fork" button. You should now be on the https://github.com/VOTRENOMDUTILISATEUR/contoso-chat page in your own GitHub account. You now have a copy (called a fork) of this lab repository in your own GitHub account! Don't hesitate to explore it, you won't break anything. CONGRATULATIONS! - You have a personal copy of the example to explore!
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: creation of the codespace
The image shows a GitHub page for the "contoso-chat" repository owned by the user "azurechatcopilot". This repository is a fork of "Azure-Samples/contoso-chat". Here are the main elements:
Main branch: "main", synchronized with "Azure-Samples/contoso-chat:main".
Folders and files: .devcontainer, .github, .vscode, data, docs, infra, etc.
Last commit author: A user named "nitya" recently added an ACA update.
On the right, there is an "About" section:
Description: The repository contains the complete process of building a RAG application with Promptly and AI Studio, including GPT 3.5 Turbo, LLM application assessments, deployment automation with AZD CLI, and GitHub actions for assessment and deployment with intent mapping for multiple LLM jobs.
At the bottom right, there is a "Codespaces" pop-up indicating that there are no codespaces for this repository and offering the option to create one on the main branch.
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: codespace preparation
Préparation du codespace
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: the codespace is ready
The image shows a code development interface, Visual Studio Code, with a README.md file open. Here's a look at what's visible:
File title: "Contoso Chat: Retail RAG Copilot with Azure AI Studio and Prompty".
Buttons: There are buttons to open GitHub workspaces and developer containers.
Table of contents of the README file:
Overview
Features - Architecture Diagram
Prerequisites
Getting Started
GitHub Codespaces
File Explorer on the left:
Multiple files and folders like .github, .vscode, dca, docs, src, .gitignore, azure.yaml, CODE_OF_CONDUCT.md, CONTRIBUTING.md, docker-compose.yml, LICENSE, and others.
Terminal at the bottom of the screen:
Command line displaying: @azurechatcaipilot ➜ /workspaces/contoso-chat (main) $
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: welcome to Microsoft Azure
We open the Azure portal: https://portal.azure.com
welcome to Azure
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: how do you want to use Azure?
The image shows a web page in Microsoft's Azure portal. The page asks the user how they plan to use Azure. There are four options for the user to choose from:
Improve data and cloud infrastructure security: This option is for those who want to use Azure to strengthen the security of their cloud-hosted systems and data.
Use platform services to build, test, and deploy applications: Ideal for developers who want to leverage Azure's platform services to manage the lifecycle of their applications.
Use AI and ML to add intelligent capabilities to applications: This option is for those who want to integrate artificial intelligence and machine learning capabilities into their applications.
Develop and deploy scalable cloud-native applications: Targeted at developers building applications that need to scale easily with demand, using Azure cloud-native services.
At the bottom of the page, there is a link to Microsoft's privacy statement and two buttons: "Skip" and "Next". The page is numbered "Page 2 of 3", indicating that this is the second step in a three-step process.
This page helps users define their goals and configure Azure based on their specific needs.
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: Azure services
Azure all services
The image shows a screenshot of the Microsoft Azure portal, specifically the "All services" section. The selected category is "AI + machine learning," which contains 22 services. Here is a detailed description of each service listed:
Azure AI Studio: An integrated environment for developing, testing, and deploying artificial intelligence models.
Azure Machine Learning: Enables you to build, train, and deploy machine learning models at scale.
AI Search: Integrates intelligent search capabilities into applications.
Azure AI services: A collection of artificial intelligence services for various applications.
Azure AI services multi-service account: Account to access multiple AI services under a single subscription.
Azure AI Video Indexer: Analyzes and indexes videos to extract useful information.
Anomaly detectors: Detects anomalies in data to identify unusual behavior.
Bot Services: Builds, tests, and deploys intelligent bots.
Computer vision: Analyzes and includes images and videos.
Content moderators: Automatically moderates content to detect inappropriate elements.
Custom vision: Creates custom vision models for specific needs.
This image is interesting because it shows the diversity of artificial intelligence and machine learning services available on Microsoft Azure, useful for developers and companies integrating AI capabilities into their applications.
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: Azure resource groups
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: the Azure deployment
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: Azure AI Studio
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: AI Studio resources
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: the Azure Ai Studio project
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: the particularities of the AI Studio project
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: the 1795 AI models available
The image shows a screenshot of the "Model catalog" page of Azure AI Studio. This page presents a list of artificial intelligence models available for various tasks. Here are the models displayed and their features:
gpt-4o-realtime-preview: Audio Generation
openai-whisper-large-v3: Speech recognition
openai-whisper-large: Speech recognition
gpt-4o: Cat Completion
gpt-35-turbo: Cat Completion
o1-preview: Chat completion
o1-mini: Cat Completion
gpt-4o-mini: Cat Completion
gpt-4o: Cat Completion
gpt-4-32k: Cat Completion
gpt-35-turbo-instruct: Cat Completion
gpt-35-turbo-16k: Cat Completion
dall-e-3: Text to image
dall-e-2: Text to image
whisper: Voice recognition
tts-hd: Text to speech
tts: Text to Speech
text-embedding-3-small: Embeddings (creating vector representations for text)
These models are used for various applications in the field of artificial intelligence, ranging from the creation of audio and visual content to the analysis of text and speech. Azure AI Studio provides an integrated environment for developing, testing, and deploying these models, making it easy to integrate advanced AI capabilities into applications.
Azure AI Studio offers a variety of AI models for different tasks, such as:
Audio Generation: For example, gpt-4o-realtime-preview.
Speech recognition: For example, openai-whisper-large-v3.
Cat completion: For example, gpt-4o and gpt-35-turbo.
Text-to-image conversion: For example, dall-e-3.
Text-to-speech conversion: For example, tts-hd and tts.
These models are used for a variety of applications, ranging from interactive content creation to complex data analysis.
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: Azure container apps
The image shows a screenshot of the Microsoft Azure portal, specifically the overview page of a container environment named "aitour-45479574-containerapps-env". Here are the visible details:
Environment name: aitour-45479574-containerapps-env
Environment Type: Consumption only
Resource group: rg-AITOUR
Location: Central France
Subscription ID: 93b4f526-6464-406e-bf46-582d467a8252
Static PI: 98.66.217.127
Applications: 1
KEDA Version: 2.15.1
Dapr version: 1.12.5
.NET Aspire Dashboard: Enabled
Tags: azd-env-name: AITOUR
On the left side of the page, navigation options include:
Overview
Activity log
Access control (IAM)
Tags
Diagnose and solve problems
Settings
Insights
Apps
Service Connector
Monitoring
Automation
Help
At the top of the page, the "Refresh" and "Delete" buttons are also visible.
Detailed explanations of each section and element:
Environment Name: Specifies the unique name of the container environment.
Environment type: "Consumption only" means that this environment is based on on-demand usage, which is often more cost-effective for fluctuating workloads.
Resource group: A logical container that groups linked resources together for simplified management.
Location: Indicates the geographic region where the environment is hosted, in this case "France Central".
Subscription ID: A unique identifier for the Azure subscription to which this environment belongs.
Static IP: A fixed IP address assigned to this environment for stable access.
Applications: The number of applications deployed in this environment.
KEDA version: Indicates the version of Kubernetes-based Event Driven Autoscaling (KEDA) used to autoscale containers based on events.
Dapr version: A version of Distributed Application Runtime (DAPR) used to simplify the creation of microservices applications.
.NET Aspire Dashboard: Indicates whether the Aspire Dashboard for .NET is enabled, allowing for better monitoring of .NET applications.
Tags: Used to organize and categorize resources for easier management and invoicing.
In sum, this page provides a comprehensive and organized view of the information that is essential for managing and deploying a container environment in Microsoft Azure.
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: the url of the container app
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: Azure container live in place
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: Azure cross-paltform CLI
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: welcome to Cosmos DB
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: what is Cosmos DB?
What is Cosmos DB?
What is Azure Cosmos DB?
Azure Cosmos DB is a fully managed NoSQL database service designed for modern application development. Here are some of its main features:
Fully managed: Azure takes care of the underlying infrastructure, allowing you to focus on developing your application without worrying about server management or scaling.
High availability: Ensures 99.999% availability, which means your applications remain accessible and reliable.
Multi-Model: Supports multiple data models such as document model, key-value model, graph, and Cassandra tables.
Low latency: Globally available with millisecond response times.
Geo-distribution: Easily distributes data across multiple Azure regions, delivering optimal performance wherever your users are.
Elastic scalability: Automatically adapts to meet user demands, whether you need to manage a small application or a large enterprise.
Security and compliance: Provides advanced security features and compliance certifications to protect your data.
Azure Cosmos DB is best suited for applications that require high availability, low latency, and elastic scalability, such as web apps, mobile apps, gaming, and IoT apps.
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: Cosmos DB Data Explorer
The image shows a screenshot of Data Explorer in the Microsoft Azure portal, for a Cosmos DB account named "cosmos-contoso-45292831". Here are the key elements:
Data Explorer: Allows you to navigate and manage data in a Cosmos DB database.
"Contoso-outdoor" container and "customers" collection**:
The "customers" collection contains JSON documents. The image shows an example of a JSON document with detailed information about a client, including:
id
firstName
lastName
address
Membership
A list of orders with order details like id, productId, quantity, date, total, shipDate, unitPrice, unitsInStock, and description.
Interface options:
Create a new container
Update, discard, delete items
"Edit Filter" button to apply filters to the displayed data
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: Azure Search Explorer
The image shows a screenshot of the Microsoft Azure Search Explorer interface. Here's a detailed description of what can be seen:
Header:
At the top of the image, there is the Microsoft Azure navigation bar with the logo and search options.
The visible URL is https://portal.azure.com/#view/Microsoft_Azure_Search/SearchExplorerRead....
Main Section:
The title of the page is "Search explorer".
The name of the selected index is "contoso-products".
There is a search bar with a "Search" button on the right.
Search Results:
The search results are displayed as JSON.
JSON starts with:
{
"@odata.context": "https://srch-45292831.search.windows.net/indexes('contoso-products')/$metadata#docs",
"@odata.count": 20,
"value": [
{
"@search.score": 1,
"id": "20",
"content": "Step into the great outdoors with the CompactCook Camping Stove, a convenient, light...",
"filepath": "compactcook-camping-stove",
"title": "CompactCook Camping Stove",
"url": "/products/ compactcook-camping-stove",
"vector": [
0.018875257,
0.003079016,
-0.0044105756,
...
]
}
]
}
Search explorer: This is a tool for browsing and querying search indexes created in Azure Search. It is used to test and validate search queries.
"contoso-products" index: An index in Azure Search is a collection of documents that you can query. Here, the "contoso-products" index contains information about various Contoso-branded products.
Search bar: Used to enter search queries. The user can enter search terms and click on the "Search" button to get results.
Search results: The results are returned as JSON. Each result contains several fields, including the search score (@search.score), the id, the content, the filepath, the title, the url and a vector.
@search.score: Indicates the relevance of the document to the search query.
id: The unique identifier of the document in the index.
content: An excerpt of the product's content.
filepath: The path of the file in the product management system.
title: The title of the product.
url: The relative URL of the product on the website.
vector: An array of numbers representing product characteristics for advanced search and classification purposes.
Using Azure Search Explorer helps to check how documents are indexed and ensure that search queries return the desired results. It's an essential tool for optimizing the search experience in applications using Azure Search.
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: Prompty
Prompty explorer tent
The image shows a code development interface, Visual Studio Code, with an open file named "chat-prompty.yml". This file contains configurations for an AI chat template. Here is a detailed description of the file's contents and interface:
File Contents chat-prompty.yml
# authors: Billet
# since: 2023
api:
kind: chat.com
operations:
- type: serverless
endpoint: https://models.inference.ai.azure.com
model: gpt-40
sample:
- input: >
{firstname: Loic Billet}
output: >
The Alpine Explorer Tent boasts a detachable divider for privacy, numerous mesh windows and adjustable vents for ventilation, and a waterproof design. It even has a built-in gear loft to store your outdoor essentials. In short, it's a blend of privacy, comfort, and convenience, making it your second home in the heart of nature!
Question: What can you tell me about your tent?
Terminal Section
Below the file, there is an output console displaying deployment logs with information messages (info) indicating the loading and calling of the GPT-40 model. Here are the transcribed logs:
[info] Loading /workspaces/contoso-chat/.env
[info] calling model gpt-40
[info] Bot setup: If you're looking for a tent that combines comfort, privacy, and practicality, the Alpine Explorer Tent might just be your new favorite outdoor companion!
Description of the special features of the Alpine Explorer tent
At the bottom of the image there is also a description of the special features of the Alpine Explorer tent:
Here's what makes it special:
- "Detachable Divider": Perfect for privacy when you're sharing the tent.
- "Ventilation": Numerous mesh windows and adjustable vents to keep you cool and comfy.
This YAML file sets up an AI chat template with the following:
Author and date: The authors and the year of creation.
API: The type and operations of the API with an endpoint and the model used (gpt-40).
Sample: Example of entry and exit for the model, describing the characteristics of a tent.
The terminal displays the deployment logs, showing that the .env file is loaded and the gpt-40 template is called. The bot release provides a detailed description of the Alpine Explorer tent, highlighting its features like a detachable divider and optimized ventilation.
AI Azure Open AI Chatbot integration with RAG CosmosDB from fork GIT POC MVP: prompty in French
It is possible to ask the model to speak about the tent in French
The chatbot describes the item in stock well and makes the customer want to buy it. Mission complete!
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: prompty answers more questions
The chatbot is comfortable talking about teleworking, coffee, IT developers...
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: AI-102 certification
After seeing this POC, a client wanted to take me on an AI project. HR asks me if I am certified as an AI-102 Azure AI Engineer, Designing and Implementing a Microsoft Azure AI Solution, it is a mandatory prerequisite to join the team. So I passed my certification:
And bam! And then it's the drama, "Fail", I miss my certif by 50 points... I am not integrated into my client's AI project... I'm so sad. Never mind, I'll continue to develop my projects in Csharp, Javascript and Python ...
While drinking my coffee to cheer me up, I reread the result of the certificate, and I see that I made a good score on the "implement natural language processing solutions" part...
And I come up with an idea for an application for social networks:
I will be able to use this Azure chatbot to answer for me on social networks while I drink my coffee. Quick writing of use cases/user stories hophop hop:
-Arrive at the coffee machine. Press the button to make the coffee flow. Meanwhile...
- Open the social network.
-As soon as I open the social network, the social network's algorithm shows me a post.
-Inject into the social network web page of the first post I see.
-Read the posts that the social network's algorithm decided to show me first.
-I create a prompt that mixes the fact of drinking my coffee with the content of the post.
-I press a button and the bot writes a response.
-I reread the bot's answer.
-I validate the answer and correct if necessary.
-I press the "published" button and it publishes the comment.
"I read the next post, and so on, until I had drunk my coffee."
Let's go! I'm going to turn these use cases/user stories into prompt and enter them into Github Copilot and it's going to create an application for me in vibecoding, tab coding...
To be continued...
AI Azure Open AI Chatbot Integration with RAG CosmosDB from fork GIT POC MVP: Need help?
Need help with Windows 11, Azure, AI, or IT in general? Fill out this form, a member of the association will get back to you as soon as possible.