- 6/2/2025
Welcome to this hands-on AI-3018 lab session, where we take a deep dive into Azure AI Foundry Service—Microsoft’s powerful AI platform for training, fine-tuning, and deploying large-scale AI models. Whether you're preparing for the Microsoft AI - 3018 Certification or looking to build AI-powered applications, this step-by-step demo will help you understand how Azure AI Foundry can be used to accelerate AI development.
🔍 What You’ll Learn in This Video:
1️⃣ Introduction to Azure AI Foundry Service: What it is & how it works
2️⃣ Setting up Azure AI Foundry in the Azure portal
3️⃣ Exploring pre-trained AI models and customizing them
4️⃣ Fine-tuning models for specific industry use cases
5️⃣ Deploying AI models for real-world applications
6️⃣ Best practices for AI model optimization & scaling
🛠️ Who Is This For?
Beginners preparing for the Microsoft AI-3018 Certification
AI & ML enthusiasts exploring Azure AI solutions
Developers, data scientists, and IT professionals working with AI model training & deployment
📌 Key Highlights:
✅ Hands-on, step-by-step demo of Azure AI Foundry
✅ How to use pre-trained AI models & customize them
✅ Fine-tuning AI models for real-world applications
✅ Best practices for deploying AI models in Azure
💡 Unlock the full potential of Azure AI Foundry and accelerate your AI journey today!
Explore Our Other Courses and Additional Resouces on: https://skilltech.club/
🔍 What You’ll Learn in This Video:
1️⃣ Introduction to Azure AI Foundry Service: What it is & how it works
2️⃣ Setting up Azure AI Foundry in the Azure portal
3️⃣ Exploring pre-trained AI models and customizing them
4️⃣ Fine-tuning models for specific industry use cases
5️⃣ Deploying AI models for real-world applications
6️⃣ Best practices for AI model optimization & scaling
🛠️ Who Is This For?
Beginners preparing for the Microsoft AI-3018 Certification
AI & ML enthusiasts exploring Azure AI solutions
Developers, data scientists, and IT professionals working with AI model training & deployment
📌 Key Highlights:
✅ Hands-on, step-by-step demo of Azure AI Foundry
✅ How to use pre-trained AI models & customize them
✅ Fine-tuning AI models for real-world applications
✅ Best practices for deploying AI models in Azure
💡 Unlock the full potential of Azure AI Foundry and accelerate your AI journey today!
Explore Our Other Courses and Additional Resouces on: https://skilltech.club/
Category
🤖
TechTranscript
00:00now with this particular note let's start our today's learning so far every time when we have
00:15a new video we always focus on latest Azure AI services and today I'm going to introduce one
00:20of the latest member of Azure AI service which is Azure AI Foundry let me just show it to you
00:27what this service is and very soon I'm going to come back with different features and
00:32characteristics with demos of Azure AI Foundry service so what are you waiting for guys let's
00:37get started oh yes and by the way I forgot to tell you that if you are liking our videos or if you
00:43want to learn AI you should subscribe to our channel but I think you've already done it if
00:47you're really serious about learning AI and making a growth in your career I think you have already
00:52subscribed thank you so much for that and let's move forward as I said Azure AI Foundry is one of
00:59the latest service of Azure AI services now not only that this is one of the service which is
01:04actually allowing you to integrate with couple of other Azure services so what is this the Azure AI
01:11Foundry service is actually having one centralized umbrella where you can actually bring all the
01:16Azure AI services together now basically they are providing something which is known as Azure AI Foundry
01:21portal which is a web portal that brings together multiple AI related services into a single unified
01:28development environment specifically this Azure AI Foundry service is going to combine multiple
01:34things first it's going to help you to deal with model catalog where not only Azure open AI and other
01:40Microsoft models are there some non Microsoft models are also available but not only Microsoft and open AI
01:47models are available but so many other elements are also available in this you can actually take care
01:52of model catalog you can actually deploy the model or you can develop AI capabilities with the help of
01:58prompt flow the generative AI models are going to help you in development testing custom data integration
02:04capabilities of Azure open AI service with this and as I said integration is also there so you can
02:11integrate as your AI vision service speech service language service or document intelligence services
02:16and your content safety services all together in one is your AI Foundry service basically as I said
02:23under this one umbrella you can actually develop and you can use whichever is your AI service you want
02:29to use if you ask me who are going to use this well I will say if you are a developer or if you're a data
02:35scientist then this service is for you now in addition to providing a unified interface for multiple AI
02:42services Azure AI Foundry also include a support for creating some new terminologies and some new entities
02:49with this like we have something called Azure AI hubs that's actually going to provide a collaborative
02:55development workspace for data scientists developers and others who wants to build AI solutions under AI hub you
03:02can create multiple projects so basically AI hub is a root level element inside one AI hub you're going
03:08to have multiple projects these projects are going to take care of your assets code and the specific
03:13solutions which you want to manage not only that you can actually take care of permissions roles and
03:18responsibilities also with your team members you're going to have scalable on-demand compute available and
03:25you can also have integration with data sources and other cloud services so storage account data lake or other
03:31configurational services you can actually integrate with Azure AI foundry service it's going to provide
03:37a web-based code development environment so you don't need to worry about installing some development
03:41environment in your computer as well as there are plenty of SDKs and command line interface libraries
03:48are available for automation all in all this is one package which Microsoft has gifted you if you are
03:54interested in AI now let me just give you a little bit introduction about AI hub an AI hub is going to
04:00provide a collaborative workspace for AI solution development and management and as you can see
04:06right now on this particular slide Azure AI hub can actually host one or more projects as you can see in
04:12the screenshot we have multiple projects under one AI hub and in the left side you can actually see we can
04:18manage quota multiple models and endpoints we can also take care of connected resources now i know you're
04:24curious about knowing this kind of things well starting from tomorrow i'm going to introduce you
04:30with couple of Azure AI foundry services so we are going to see each and every feature which is provided
04:36by Azure AI foundry service now if your question is Maruti what exactly we are going to plan how exactly
04:42we are going to learn it what kind of demos are there and what kind of topics we are going to learn
04:46in the coming week well the answer is here when we have to use AI foundry this question is actually
04:54going to be giving you what exactly we are going to learn so as you can see here on the slide
04:59Azure AI foundry is providing everything developers and data scientists need to develop agents and
05:05cutting-edge market-ready responsible generative AI applications if you have ever dream of developing
05:12your own generative AI application or you're working in a project in your organization where you are
05:17using generative AI application this is for you it's a go-to platform when you need to do multiple
05:24things you can create and manage multiple AI projects you can develop generative AI applications
05:30you can explore available AI models and as i told you you have multiple models lot and lot models are
05:37available you are not limited to only GPT models you can monitor and evaluate AI models you can
05:43integrate with Azure services and you can build responsibly because it is actually compliant with
05:49the responsible AI guidelines not only that one of the most favorite topics of my clients when i deliver
05:55my corporate trainings for my corporate clients they are very interested in one thing which is known as
06:00RAG retrieval augmented generation and that's also something which you can use inside
06:06Azure AI foundry all in all this is going to be a very important and interesting service for
06:12all of us and from tomorrow onwards we are going to pick one by one capability of Azure AI foundry service
06:19and i'm going to introduce it to you so stay tuned it's not done it's just started and if you have
06:26not used Azure AI services so far you're not late this is a perfect time before i start this thing
06:32i strongly recommend you to go through other Azure AI services like language service vision service or
06:39document intelligence service even if you are familiar with Azure open AI service that is perfect
06:44now these are all services which are combined inside one separate centralized portal and that portal is
06:50actually known as Azure AI foundry portal and i will try my best to give you an overview of Azure AI
06:56foundry portal but this is one of the huge service so it's not possible to cover this thing in one single
07:02video but we'll try to use it today as much as possible and obviously like always we are going
07:07to learn step by step so without wasting time let's get started with Azure AI foundry the first thing
07:15which you can do right now is you can go to ai.azure.com this is a url which is giving you a portal of
07:22Azure AI foundry let me tell you one thing Azure AI foundry portal is not having any particular cost
07:28associated with that you can give it a try but inside this Azure AI foundry portal if you create
07:34any kind of a model deployment or if you're going to use any of the underlying Azure services with that
07:39those Azure services and model deployment will be chargeable so basically portal is free but whatever
07:45service you're going to use inside that will be chargeable i am inside my Azure AI foundry portal i
07:51have already logged in with my account here and that's the reason i'm getting this page which is
07:55showing me welcome to Azure AI foundry maruti thank you now let's move forward to creating a
08:02new project let me tell you one thing guys those who are new to Azure cloud this is not Azure cloud
08:07portal Azure cloud portal you can access through this which is portal.azure.com and you can see right
08:13now this is a separate portal where you can create Azure cloud resources i am first going to deal with
08:19Azure AI foundry portal and we will try to create a project here based on this whatever resources you're going
08:25to need those underlying services and resources will be created on your Azure cloud portal so basically
08:31in order to use those services you need to have an Azure subscription also now it's giving me a project
08:38name so i'm trying to create a new project in this case whenever you create a project it will be created
08:43under something called hub so we have something called AI hub and inside one AI hub you're going to get
08:49multiple projects or you can create multiple projects now i am going to give some kind of a meaningful
08:54name for this so it's giving me maruti proj and i'm giving some random number at the end of this
09:00which is something is there it's also showing me that when you create a project you are going to use
09:06this subscription it will be created under this resource group and this location as well as it's
09:11going to create certain new things for me like it's going to create a new AI hub this new AI hub is actually
09:17going to be created first and inside that this project will be created later on if you want to use
09:24same hub then you can do this and that's the reason i'm giving some meaningful name maruti hub with
09:30some number at the end of this uh resource group name is fine this has to be unique so that you should
09:36not get confused with that so i'm okay with that location is also fine east us too this is going to
09:42connect with azure ai service or azure open ai service in this now you can see they are automatically
09:47going to create a new azure ai service with that and the name of the service is this now if you want
09:53to connect with other services later on you can do this thing and i'm going to show you how to do
09:58that thing today but before we do that i think everything is fine do i need to connect with azure
10:03ai search as of now well today's demo is not going to associate with azure ai search so i do not want
10:08this i'm going to click on next it's giving me the review perfect i'm going to click on create and now
10:16this process is actually going to take a little time it can take two to five minutes of time depends upon
10:21the selection of services which you have done as of now it's showing me all the things are under
10:26provisioning so i have to wait it's validating my resources and once the resource validation is done
10:31it's going to start the deployment while we are waiting for this deployment i should remind you one
10:36thing that if you're liking our videos please press the like button hit that bell icon and please
10:42subscribe to our channel this is really important for us and your support is always going to motivate us
10:47also i want to remind you if you are looking for any specific azure cloud or ai related topics in
10:53our videos please put that thing in the comment of this particular video we will surely create a video
10:58on that particular topic mentioning your name in that that this person has actually suggested that
11:03thank you so much okay my project is successfully created as you can see here and obviously the
11:10underlying services are also created now first thing is right now you can see in this particular portal that
11:16we are on the overview tab and it's showing me that there are some endpoints and keys which are
11:21available associated with my azure open ai service so basically this azure open ai service was part of
11:27this process and it got provision when we created the new project also at the right side you have some
11:32project details which will be useful later on and if i scroll down it's showing me some kind of
11:37steps like first you can define and explore you can build and customize and then you can assess and
11:43improve now i do not want to go into this steps right now in the left side section we have playgrounds
11:48we have ai services we have model catalog now there are two interesting things which i want you to see
11:53today in this particular video the first thing is i want to click on model catalog when i do this thing
11:59this model catalog is actually going to show you a list of llms large language models not only from microsoft
12:07it is actually including open ai hugging face even meta models are also part of this now you can see
12:13right now we have a hugging phase we have o3 mini we have deep seek r1 we have four o's audio preview so
12:21even the latest deep seek is also part of this thing if you are looking for a specific model for
12:26your project i'm sure you will find a related model in this particular one obviously this models are trained
12:32on various number of parameters and based on those number of parameters on which it got trained the
12:38performance the capability and the costing of this models are going to be very so obviously if you want
12:44a specific video on highlighting certain characteristics of all of these models i will create a separate video
12:50on that in the coming modules but as of now this model catalog is actually giving you a lot many models
12:57inside this compared to a traditional azure open ai service now the second thing which i am really
13:03interested right now is this section which is management center in the left bottom corner if i
13:08click on management center it will take you to this particular page and let me tell you this is very
13:13important page for you this is going to show you your specific connected resources with this particular one
13:20so you can see right now it's showing me that azure open ai service azure ai service azure blob storage
13:26this kind of things are actually connected with my project which i have created with this number 1861.
13:33it's also showing me that i have a section here which is called all hubs and projects when i click
13:39on this it's going to show me whatever projects and hubs i have created now obviously i have one hub which
13:45is this one and then this is my project which i have created which is under that hub you can have
13:50multiple hubs you can have multiple projects in each hub and that's how you can configure that why
13:55exactly we are going to create multiple projects well the obvious answer is maybe in our organization
14:00we are going to have multiple teams who are going to work in this project and yes your azure specific
14:05role based access control permissions everything is available here next you can take care of the
14:11quota obviously and then you can take care of this section where you can control the users
14:16and permissions as of now i am the only person who is actually an owner of this particular project
14:22as well as my project is also having a role associated with ai administrator role in that
14:28because my project is going to have capabilities to deal with my azure ai resources which are connected
14:33with that project now i am going to focus on role-based access control separately in another video
14:39but right now this is available here then you have models plus points we know this thing that every
14:45time when you try to deploy a model it's going to be available with the specific quota with the specific
14:50configuration and then with the help of endpoint you'll be able to use it now all those model
14:55deployments you'll be able to do it here you can deploy a base model or you can deploy a fine-tune
15:00model also now after this last and the important thing which i want you to focus today is this
15:07connected resources as we know there are multiple services which got created with the help of this
15:12project and ai hub these resources are actually created in your azure portal so basically where exactly this
15:19is created the answer is you have to go to your azure portal you have to go to resource group section
15:27inside that you're going to find a resource group which is having a name which got created
15:33automatically with that process and now you can see this is my resource group which is actually
15:38showing me that i have four five services which got created with this you have a key vault you have
15:43azure ai service azure ai hub azure ai project and a storage account so these things are actually got
15:50created with this now can i create more than one azure ai service or azure open ai services with my
15:57ai project answer is yes you can do that thing let's try certain things with that so i'm going back to
16:03my connected services section here which is showing me that some of the services are there but i already
16:08have my one azure open ai service now now suppose that in this particular ai project you maybe need
16:16another azure ai service or azure open ai service which is maybe belongs to a different region and
16:21different location in any project it is very common that you have customers across the globe and there are
16:27chances that only this east us related services are not enough maybe i need another service in the
16:33different region all together well if that is the case you have to go to your azure portal first i'm
16:39here i'm going to create a new resource i'm going to search for azure ai services so i'm going to search
16:47for azure ai service when i search for this remember you are going to choose azure ai service
16:57so i have this azure ai service style available here i'm going to click on create azure ai service
17:04this time while creating this i am going to choose same resource group but i'm going to change the
17:09region so i do not want any u.s specific region this time i'm looking for a region which will be
17:14somewhere in europe so i'm choosing uk south the name of the service i'm giving maruti second ai service
17:26pricing tier we are okay with standard and obviously content review policy we always agree with that i'm going
17:33to click on review plus create now this service we are provisioning with the help of azure portal and
17:40obviously once this is successfully provisioned the next step is we want to connect this in that same ai
17:45project under that ai hub so i'm just creating this
17:51okay this deployment is successfully completed let me click on go to resource to just check whether my
17:57service is successfully provisioned or not it's showing me status is active location is uk south
18:03and that means now i can go to my project i'm inside my management center i'm going to click on
18:10to click on connected resources and now this is a place where we have four resources which are
18:14connected i'm going to click on new connection i'll choose azure ai service remember you can choose
18:22from any of the services which you have already provisioned uh you have ai specific service you
18:27have data specific service you also have indexes and other resource type which are available i'm going
18:33to choose right now azure ai service because this is what we have created they have options like browse
18:39resource or you can enter it manually now in my case i am actually able to see that there is one
18:45resource which is east us which is there and there is another resource which is uk south now you can see
18:50east us one is the one which is already connected with this so i don't need to add connection but this is
18:55the one which is my second service which i've created is actually the one so i'm going to click
19:00on add connection on this wait for some time and you can see it's showing me connected with the green
19:07tick which means that this service is actually successfully connected with this i'm going to click
19:12on close and it's going to refresh this list and yes you can see right now my new service is also part
19:18of this list my new service which is a second ai service is actually created here
19:24and it's also available here so these are the two services which got connected with my project
19:29and these are the endpoints and keys which are available with that i'm happy with this
19:34i can connect multiple services even from the different locations into one single ai project
19:40cool okay now once we are good with this let's go back to our azure ai foundry portal
19:45i do not want to do anything in the management portal now i'm going inside my project and in this
19:50project the next thing which we want to explore is ai services let's see what kind of ai services are
19:55actually available inside this as you can check here all your ai services like speech service language
20:02service plus translator vision service plus document intelligence service and the content safety all are
20:09part of your package so you can use any of the services in your project i also want to remind you that if
20:17you're not familiar with any of the services we have already covered this as a part of our azure ai
20:22fundamental course ai 900 and all these videos are available in the same channel so if you have
20:28missed learning speech service or vision service or document intelligence service i strongly recommend
20:34you to check out these videos which are available in the list of azure ai 900 fundamentals now right now
20:41let's try language plus translator it's showing me a language playground i can try the language playground
20:48if i want or if i scroll down it's showing me a sections like what's new from language what kind of
20:54things we can do with this now there are so many things which i can try it out right now
21:04i'm going to try some of the translations so i'm not going with the classifying text or
21:08summarization i'm clicking on the translation tab and inside the translation we have document
21:14translation and we have text translation let's say as of now i want to try a text translation demo to
21:20check whether it's working or not it's giving me a page where i can see the code if i want or it's
21:26showing me a text and the translation which i can associate here now i can use any of the samples which
21:32are available here but in this case i want to try with my own text so i'm going to click on this tab
21:37which is try with your own remember guys this particular portal is having so many things inside
21:42that whenever you're using this portal first time there are chances that you're not very comfortable
21:47with this kind of a ui but i request you to please spend some more time with this and you're comfortable
21:52that where to click what to do inside that and all so just spend some more time and then you'll get
21:58it i'm just clicking on this try with your own and it's showing me that what will be your source
22:04language what will be your target language now source language it can have an auto detect option so it's
22:09going to detect whatever i'm going to type inside that and the target language i can choose let's say
22:14i'm going to choose that my target language is going to be there are plenty of languages let's say i'm going
22:21to select russia and let me put something like uh hey friends i am really
22:42enjoying enjoying this learning let me just click on translate
22:57and of course i don't know how to read russian but yeah this is maybe converted into a russian
23:03let me change to some other language which maybe i can read and i can understand so let me just go into
23:09my local language which is hindi and i'm going to click on translate
23:15now he's saying yes this is something which i can read because i'm from india so namaskar
23:23now this is something which is uh no proper translation into hindi now let's say
23:30we are putting some word is your ai it's amazing and
23:41i will use it in my work
23:46of course this will sound little weird when you see this thing in hindi but yeah let's try
23:50yeah it's just understanding the thing so is your ai they are not trying to convert that into hindi
23:55it's saying azure ai adbuth hai aur mein isse apne kaam mein upio karo ga let me see something which is
24:04we're going to try it into french and yes it's translated into french so it's how easy it is right
24:12now now obviously this is something if you're a business user you're going to try the same but if
24:16you're a developer i strongly recommend you try the samples which are available here so you can use this
24:22english to arabic kind of a samples which are available you can click on the view code for any
24:27of this thing and it's actually going to show you a code in multiple languages it's going to show you
24:31a code in python c sharp javascript this code is a working code you can directly copy paste and you can
24:38put it into your whichever developer environment you're using and you can try those samples with the
24:42resource key as easy like that now i strongly recommend you to try this thing and if you're trying and
24:50if you're liking or not liking this please put that thing in the comment and if you have any questions
24:54also feel free to ask us we are here to help you on that this is cool this is good now if we are dealing
25:03with generative ais obviously the first thing which we need is we need our own model so we have to deploy
25:09a model and that is what we are going to do next right now so now if you see in the left side section
25:15we have my asset section in which i have model plus endpoint i'm going to click on this model
25:21plus endpoint and this is a place where i can deploy a new model i'm going to click on deploy
25:26a model and let's say we are first going to try to deploy a base model right now so i'm choosing
25:31deploy base model it's going to show me the list of the models which are available let's say as of now
25:38i'm going to use gpt4 so i'm just searching for gpt4 here it's showing me gpt4o gpt4o mini i'm going
25:49to choose gpt4 as of now and this is a model which i want to use so i'm going to click on confirm
25:59now because we are trying to deploy this model right now it's going to ask us a configuration that
26:04what kind of deployment type you want do you want a global standard standard global batch now these
26:09are different deployment types which we are going to see later on in depth because this is the first
26:13introductory video i'm not going with this kind of a depth right now i am going to choose right
26:18now that my deployment type is going to be standard deployment and that's why somewhere in the deployment
26:23details i'm going to specify some configurations my connected service will be this one uh tokens per
26:30minute rate i want to reduce to 5k which is fine and the content filter i think its default is fine
26:37dynamic quota i do not want so i'm going to disable this so this is what a configuration i'm looking
26:41for for the gpt4 i'm going to click on deploy yes my gpt4 model is successfully deployed trust me guys
26:51i have worked with the other azure ai portals also earlier even azure openai studio was also there
26:57this particular portal is much much much better than that it's fast it's accurate it's easy to
27:03understand and the user interface is super neat and clean i'm sure you agree with me on this now let's
27:09start this thing with the model so if you see my model is successfully deployed with my configuration
27:14which i have done first thing first i want to try this thing in the playground so i'm going to click on
27:19open in playground this is going to associate with the chat playground because we have gpt4 model which is
27:24a perfect model for chat i hope you know some basics of chat playground configuration this chat
27:30playground is actually having three sections we have a section which is a setup section where you're
27:35going to choose your deployment you're going to choose what kind of assistant you want to configure
27:39for this and then the right side you have your chat box so basically this is where you're going to
27:44type your query which we call prompt and then based on that is going to give you a response which
27:49we actually call completion so yes we are basically going to do request response but here we call this
27:55thing prompt and completion now when we start typing with the sample prompt there are some sample prompts
28:01also which are available but i'm going to try my own so right now i want you to see here that is showing
28:06you that you are an ai assistant that helps people in finding information so i'm just putting a prompt right
28:12now like this how can i use azure ai services in a software development project now let's say i don't
28:19know how to do this thing so i'm just going to put this thing in the prompt when i do this let's see
28:23what this azure ai assistant is actually going to give me an answer it's saying azure ai service can
28:29be used in the software development project in various ways the first one is natural language processing
28:35and it's giving me some information about that you can use text analysis api luis which is a language
28:41understanding intelligence service bing spelling check computer vision service speech recognition
28:47service so so it's actually listing out all those services which i can use and then it's going to give
28:51me that thing now right now the generation of this thing is little slow uh that's because i have done
28:56some configurations in my azure ai foundry but for you maybe this is going to be much quicker when you
29:01do that thing um yeah we can knowledge mapping we can do bot development we can do machine learning
29:07all the services are actually part of azure ai services and then
29:11and we can use this thing while developing projects for our customers and client
29:15fine i'm happy with this response now let me try a few more things because this is an ai assistant
29:20right now maybe it cannot answer all kind of questions so i'm going to change something in the
29:25left side section where we have ai assistant i'm going to say you are not an ai assistant you are
29:30a development assistant so you are a developer
29:39assistant
29:44you're a developer assistant
29:49that helps python developer so i'm saying that helps python
29:54developers
29:58to write
30:01logical code so i do not want someone who can give me an advice and information only i want an assistant
30:08who can actually help me to write a python code so i'm just going to say you are a developer assistant
30:13and then once you change this kind of a system message you just click on apply
30:17and continue once you apply these changes this will be saved in this particular model gpt4 model
30:24and let me put a question which is associated with this so i'm going to say
30:30i want to write
30:34fibonacci now intentionally when i'm adding this thing i'm saying fibonacci with this wrong spelling so i know
30:40the right spelling is this but intentionally i'm adding a wrong spelling just to check whether this is
30:44able to understand or not so i'm just saying i want to write a fibonacci logic
30:51can you help me with that
30:54also notice uh i am not mentioning that i want to do this thing in python but i just mentioned here
31:00that okay you're a assistant for a python developer so they are going to assume that thing
31:04and i'm just saying can you help me with that and i'm going to send this prompt
31:08so i want to write a fibonacci logic fibonacci spelling is also wrong but it's going to
31:13identify that thing automatically so let's see that thing what is a response which they are giving
31:17with this yes i got a response it's saying sure i can help with that here is a simple implementation
31:23of fibonacci sequence and then it's showing me a function in the python which is having a function
31:29name some parameter and the logic and also after this is showing me that explanation of this particular
31:36function also that what this is actually doing this is cool this is actually generating the code and
31:42it's showing me what exactly that code is doing now let me add something more i'm saying uh can you
31:50write some unit test
31:54for above code
31:59so i generated a code with that one prompt and now with the second prompt i'm saying i want to write a
32:04unit test case also i'm a good developer so i do not want to skip this part i am saying can you
32:10write a unit test for the above code they're saying sure here are some of the sample unit test for
32:15fibonacci function which is python built and it's showing me importing the unit test package library here
32:22it's showing me the function and very soon we are going to have the unit test also for this function
32:31as you can see right now it's generated it's showing me that class which is having taste fibonacci and
32:36then it's showing me the parameters associated with that they have created the unit test association
32:41with this and then they're also configuring that as well as in the below section they are now
32:46explaining what exactly they have done in the unit test cases amazing now this kind of configuration is
32:53actually very important when you are generating a code so basically this is your section where you can
32:58actually change how exactly you want your assistant should behave it's going to be an ai assistant it's going
33:04to be a developer assistant you also have some kind of a prompt samples which are available here
33:08so if i click on the prompt sample they are going to show me that maybe we have a prompt sample which
33:13is going to help me in marketing writing assistant or maybe i want a shakespeare writing assistant where
33:19the assistant is going to type in a shakespeare style maybe i am an author and i want to publish a
33:24novel on that so maybe i can take the same kind of a writing pattern or maybe i can have a json formatter
33:30assistant which is going to help me to take care of the indentation and formatting in the json you
33:35can use this particular chat playground as you want and then you can customize and you can see how
33:41exactly your assistant is working there are a lot many things which you can do inside this i'm not going
33:46with all the things right now if you want a detailed video of the chat playground with gpt4 you can just
33:51comment down in that particular video and i will take you through all this part i will take you through
33:56the parameter section where changing the parameter values is going to change the responses here i will
34:02also take you to something which is known as few short learning so we have actually some kind of a
34:08sections where we can add some examples we can add some variables and using that we can actually
34:13customize our prompt but as of now i think this is it remember one thing whenever you're using any of
34:19the azure ai services in your own subscription after using this after trying all these particular services
34:26i strongly recommend that you go back to your subscription in the azure portal and you make
34:31sure that whatever services and resources which you have created you're going to delete it so this
34:36is my resource group in which i have created everything and i do not want to waste my money so
34:41all i'm going to do is once i'm done with something i'm always going to delete the resource group remember
34:46this is very important do not keep all the resources there otherwise it's going to cost you every minute
34:51so it's very important you're going to go and delete your resource group which i'm doing right now
35:02and once my resource group is deleted my costing will be stopped remember deletion is also going
35:08to take some time so same like your creation of the service is taking time deletion is also going to
35:13take a time because this is an irreversible process once you delete the resource you cannot get
35:18that thing back so i started by creating this thing with azure ai foundry portal and azure portal
35:24and now we are just deleting this thing from azure portal all the resources and services will be
35:28deleted anyhow azure ai foundry portal is not going to cost you anything so if these resources are
35:33deleted you're good to go that's it for today thank you so much guys i'll see you tomorrow with the
35:38another dose of azure ai services see you tomorrow
Recommended
28:43
|
Up next