I just started checking Azure Ai, it would be nice to have a video with a super mini project that would actually do something, I don't care if it will be in Python or JS :)
@@AdrianTwarog Awesome! I am totally looking forward to this as well. I am not an experienced programmer but for my bachelor thesis I have to try to build a Microsoft AI Search RAG application that can help generate text blocks & maybe even (simple) process diagrams from exisiting quotas from similar projects (IT consulting). So it needs to find the appropriate PowerPoint, PDF, Word and maybe Excel documents, understand what is relevant and generate relevant content. In the best case I can somehow define a quota form where it would input the generated content in the right format.
Thanks for the info, "but"... Your efficiency should not be measured along the shortest time you take to say all that you need to say, but by the average time it takes for the listeners to absorb the content of what you're saying (among other factors). We know that you know what you are presenting, and that's not the goal (especially when you try something which requires another region, a problem you toss away by saying "use another _[I don't remember what you said]"_ ) ... ;-) I barely had time to click for a pause on specifics that were shown for split seconds ! lol One trick is to project yourself as if you were in a classroom with students, rather than a camera: when you explain a specific process, the student must have retained what you are saying about it from the beginning at least up to the end of its presentation, in order to understand it. Maybe not to memorize it, but at least to have a good idea about it and what it implies if you refer to it later on (as there are many dependencies). When the train passes by, it must be slow enough so that we're able to count its wagons, not just see the train as one whole emotional "TGV" in a hurry to reach the station (we're not tik-tok cows lol).
@@ywueeee From the Merriam-Webster, the meaning of CONCISE is marked by brevity of expression or statement : free from all elaboration and superfluous detail. That's not what I'm talking about.. It's not like he has to pay x$ per xx seconds, I don't understand why he forces (to himself and his viewers) such a pace.
@@HawkFest he has to! editors charge hourly? also there's no point of going slow just for the sake of ad revenue when he can pack all the information needed in a short time, the viewer can select their own pace by pausing and replaying. people have low attention span nowadays better to keep it short and quick than lose people from clicking on the video because it's too long..
@@ywueeee It's not about "going slow" for ad revenues, it's about going at a *_standard_* pace for the sake of viewer's comprehension : for the understanding of a subject that is not achieved by listening to words in a sentence, but to sentences in a coherent whole that delivers knowledge. On top of merely understanding a language, we also need the brain to assemble/create knowledge out of this. Now, regarding your argument resolving into not leaving enough room to advertisers (that's a weird twist, the goal becoming a race vs advertisers LMAO), you also talked about his expenses e.g. editors charging hourly. But what about his revenues then, doesn't he have such right, especially when we can enjoy his efforts for free? I wouldn't see anything wrong with this, but that's not the point.. More over, if we invariably need to edit the end result, then his "editor" is doing a very bad job and shouldn't get paid more that at the minimal rate, if at all. Yes we can just change the pace to "slow-mo", click every 10-15 s. to pause the clip, go backward/forward etc. But being obliged to break and slice and continuously edit the presentation in such way, is not a recipe for a good and efficient listening, as it's an added effort distracting from the effort to understand (more over when it's a knowledge that you must edit at the same time that you are acquiring it: you'd be in the worst position to do this): I'd prefer to read an article or a reference manual.
8:11 Pls you need to help me everytime i add a data source it says inefficienct funs, so it instanty uses the 1k tpm or whatever its called so its impossible to search anything
Question, would you recomend deploying a web based app through azure app services? or host the front end HTML on a website using WP and create your own plugin?
wow exactly what I was needed, thank you. I'm trying to build an AI supported movie recommender service that will recommend movies based on a selected movie. let's see if Azure AI is the way to go
So what AI service is this using in the background when you upload files and data ? Just trying to map out the cost of running this for couple of coding books.
there is a checkbox during index creation which you can set on and off which enables vector search, you swill also need a deployment for embeddings like ada-text-embedding-002
Anybody knows how to solve this error Validation error at #/data_sources/0/azure_search/parameters/index_name: Input should be a valid string. It happens after uploading my pdf on azure ai studio.
For some reason i dont have the {x} Add variable button in my "Variables" menu:th-cam.com/video/fQ9RFR1KTbY/w-d-xo.html Is is broken in Azure AI Studio right now or do i need to turn it on somewhere?
Thanks for the informative video! I find the zooming in blurry and has room for improvement. Maybe a faster zooming animation is better? It's hard to read during the animation.
I'm hoping my company is about to look at Azure (and Fabric/Dataverse) for our AI and data solutions. I know this is more of a coding channel but any video related to helping us help our end users and business would be great. I don't know what is possible but we've toyed with the idea of a customised company chatbot on SharePoint but also AI apps to help desktop users with the apps we use. The main thing we'll likely use AI for is data analytics from our various sources (external and internal).
The project failed in the end. You gave some excuse to token issue. Is this what I can expect after I've purchased your paid materials'?. Try making an update to this video to show how Completion works. I'd feel more confident in buying your materials then.
It seems that the Azure OpenAI services are locked behind a subscription that requires enterprise enrollment. I was not able to enroll with a personal email address.
Love the content but I'm struggling with the delivery. A slight pause between sentences would make this a 5 star video. As is, it is hard to follow. listening.
@@AdrianTwarog your efficiency should not be measured along the shortest time you take to say all that you need to say, but by the _average_ time it takes for the listeners to absorb what you say (among other factors). We know that you know what you are presenting, and that's not the goal... ;-) I didn't even have time to click for a pause on specifics that were shown for split seconds ! lol One trick is to project yourself as if you were in a classroom with students, rather than a camera: when you explain a specific process, the student must have retained what you were saying about it from the beginning at least up to the end of its presentation, in order to understand it. Maybe not to memorize it, but at least to have a good idea about it and what it implies if you refer to it later on (as there are _many_ dependencies). When the train passes by, it must be slow enough so that we're able to count its wagon, not just see the train as one whole unit (we're not cows lol).
Setting up private ML Services through Terraform. I'm currently having issues with the compute quotas since they already got assigned but for some reason is not working.
@@AdrianTwarog hi,thanks for reply 🥹yeah, passing is a little distracting, I think it is good for entertaining videos, but less helpful for educational,when you want to focus on content more then on changing frames. I watched 2 minutes and could not recall information, because I was focusing on changing animation so much 🤷♂️
@@Sotov I don't know what you mean with "distracting passing" and "changing animation", I rewatched the video and can not find any unnecessary or distracting animation BUT yes the PACING IS TOO FAST! I had to slow down to 0.75 (which sounds a bit awkward) and I usually watch videos at 1.25 or even 1.5.
@Adrian my scenario is like below : i have x-train excel file record : ID Month Year Brand Platform Agency 110 1 2025 pepsi Google Starcom 100 1 2025 pepsi Google Starcom 101 2 2025 pepsi Google Starcom 102 1 2025 CocaCola Facebook Starcom i have y-train excel file record : (for find) ID Month Year Brand Platform Agency 102 2 2025 pepsi Google Starcom i want use azure ai to upload both file and need result accuracy match with percentage or nearby record
absolute stunning video, quick, clear, concise, just what was needed to help me get up and running.
Your videos have a distinctive quality aspect, Thanks a lot !!
I just started checking Azure Ai, it would be nice to have a video with a super mini project that would actually do something, I don't care if it will be in Python or JS :)
I have that coming soon!
@@AdrianTwarog Awesome! I am totally looking forward to this as well. I am not an experienced programmer but for my bachelor thesis I have to try to build a Microsoft AI Search RAG application that can help generate text blocks & maybe even (simple) process diagrams from exisiting quotas from similar projects (IT consulting). So it needs to find the appropriate PowerPoint, PDF, Word and maybe Excel documents, understand what is relevant and generate relevant content. In the best case I can somehow define a quota form where it would input the generated content in the right format.
@@AdrianTwarog is that available now by any chance? thanks
@Adrian, please upload the next part.
Informative video. Thank you! (Commenting this at 4:10)
Great quality video, thank you for creating and sharing it :)
Thanks for the info, "but"... Your efficiency should not be measured along the shortest time you take to say all that you need to say, but by the average time it takes for the listeners to absorb the content of what you're saying (among other factors). We know that you know what you are presenting, and that's not the goal (especially when you try something which requires another region, a problem you toss away by saying "use another _[I don't remember what you said]"_ ) ... ;-) I barely had time to click for a pause on specifics that were shown for split seconds ! lol
One trick is to project yourself as if you were in a classroom with students, rather than a camera: when you explain a specific process, the student must have retained what you are saying about it from the beginning at least up to the end of its presentation, in order to understand it. Maybe not to memorize it, but at least to have a good idea about it and what it implies if you refer to it later on (as there are many dependencies). When the train passes by, it must be slow enough so that we're able to count its wagons, not just see the train as one whole emotional "TGV" in a hurry to reach the station (we're not tik-tok cows lol).
he's trying to keep it concise, while you have the option to pause, rewind and playback
@@ywueeee From the Merriam-Webster, the meaning of CONCISE is marked by brevity of expression or statement : free from all elaboration and superfluous detail. That's not what I'm talking about.. It's not like he has to pay x$ per xx seconds, I don't understand why he forces (to himself and his viewers) such a pace.
@@HawkFest he has to! editors charge hourly? also there's no point of going slow just for the sake of ad revenue when he can pack all the information needed in a short time, the viewer can select their own pace by pausing and replaying. people have low attention span nowadays better to keep it short and quick than lose people from clicking on the video because it's too long..
@@ywueeee It's not about "going slow" for ad revenues, it's about going at a *_standard_* pace for the sake of viewer's comprehension : for the understanding of a subject that is not achieved by listening to words in a sentence, but to sentences in a coherent whole that delivers knowledge. On top of merely understanding a language, we also need the brain to assemble/create knowledge out of this.
Now, regarding your argument resolving into not leaving enough room to advertisers (that's a weird twist, the goal becoming a race vs advertisers LMAO), you also talked about his expenses e.g. editors charging hourly. But what about his revenues then, doesn't he have such right, especially when we can enjoy his efforts for free? I wouldn't see anything wrong with this, but that's not the point.. More over, if we invariably need to edit the end result, then his "editor" is doing a very bad job and shouldn't get paid more that at the minimal rate, if at all.
Yes we can just change the pace to "slow-mo", click every 10-15 s. to pause the clip, go backward/forward etc. But being obliged to break and slice and continuously edit the presentation in such way, is not a recipe for a good and efficient listening, as it's an added effort distracting from the effort to understand (more over when it's a knowledge that you must edit at the same time that you are acquiring it: you'd be in the worst position to do this): I'd prefer to read an article or a reference manual.
8:11 Pls you need to help me everytime i add a data source it says inefficienct funs, so it instanty uses the 1k tpm or whatever its called so its impossible to search anything
Question, would you recomend deploying a web based app through azure app services? or host the front end HTML on a website using WP and create your own plugin?
Thank you man 🙏🏻 I learnt a lot!
wow exactly what I was needed, thank you. I'm trying to build an AI supported movie recommender service that will recommend movies based on a selected movie. let's see if Azure AI is the way to go
Oh that’s a pretty good idea!
So what AI service is this using in the background when you upload files and data ? Just trying to map out the cost of running this for couple of coding books.
Can you do video on azure ai vision service and custom object-detection / segmentation with the code utilizing azure ai studio ? Thanks in advance
when i start this on azure ai studio , when i go to create a new project i get hub insted of resource , what to do plz help
does azure ai studio come with its own webapp deployment for the chat model as well?
How do I tell the api to only reply to prompts based on the data I have uploaded and indexed?
Thank you fren :)
i want to use the azure studio api to do the document upload and indexing
What drawing tool are you using ?
How to switch your mouse to pen smoothly? Its interesting
When we upload documents to AI Blob, does it perform embeddings and utilize vector database behind the scenes.
there is a checkbox during index creation which you can set on and off which enables vector search, you swill also need a deployment for embeddings like ada-text-embedding-002
love it!
running like you have to catch the train
He forgot to pee before making that take
Anybody knows how to solve this error Validation error at #/data_sources/0/azure_search/parameters/index_name: Input should be a valid string. It happens after uploading my pdf on azure ai studio.
For some reason i dont have the {x} Add variable button in my "Variables" menu:th-cam.com/video/fQ9RFR1KTbY/w-d-xo.html
Is is broken in Azure AI Studio right now or do i need to turn it on somewhere?
It would be so helpful to see how to create a chat bot from an URL for all my company to use.
Hy. How could I contact you privately ?
Thanks for the informative video! I find the zooming in blurry and has room for improvement. Maybe a faster zooming animation is better? It's hard to read during the animation.
I'm hoping my company is about to look at Azure (and Fabric/Dataverse) for our AI and data solutions. I know this is more of a coding channel but any video related to helping us help our end users and business would be great. I don't know what is possible but we've toyed with the idea of a customised company chatbot on SharePoint but also AI apps to help desktop users with the apps we use. The main thing we'll likely use AI for is data analytics from our various sources (external and internal).
Create video for startup in Azure AI studio :)
The project failed in the end. You gave some excuse to token issue. Is this what I can expect after I've purchased your paid materials'?. Try making an update to this video to show how Completion works. I'd feel more confident in buying your materials then.
Thanks a lot for sharing. How to create an interface on Microsoft Teams?
It seems that the Azure OpenAI services are locked behind a subscription that requires enterprise enrollment. I was not able to enroll with a personal email address.
Love the content but I'm struggling with the delivery. A slight pause between sentences would make this a 5 star video. As is, it is hard to follow. listening.
Notes for future videos, thanks for the feedback!
Trying reducing the video speed. The voice sound a bit weird but it’ll give you time to think :)
@@AdrianTwarog your efficiency should not be measured along the shortest time you take to say all that you need to say, but by the _average_ time it takes for the listeners to absorb what you say (among other factors). We know that you know what you are presenting, and that's not the goal... ;-) I didn't even have time to click for a pause on specifics that were shown for split seconds ! lol
One trick is to project yourself as if you were in a classroom with students, rather than a camera: when you explain a specific process, the student must have retained what you were saying about it from the beginning at least up to the end of its presentation, in order to understand it. Maybe not to memorize it, but at least to have a good idea about it and what it implies if you refer to it later on (as there are _many_ dependencies). When the train passes by, it must be slow enough so that we're able to count its wagon, not just see the train as one whole unit (we're not cows lol).
Setting up private ML Services through Terraform. I'm currently having issues with the compute quotas since they already got assigned but for some reason is not working.
What is wrong with edit? Feels like I am watching a long tik tok 😢
New editing team I’m trying for my videos… I take it the pacing is too fast?
@@AdrianTwarog hi,thanks for reply 🥹yeah, passing is a little distracting, I think it is good for entertaining videos, but less helpful for educational,when you want to focus on content more then on changing frames. I watched 2 minutes and could not recall information, because I was focusing on changing animation so much 🤷♂️
@@Sotov keep calm and slow down the vídeo. Too much information, but I liked it!
@@Sotov I don't know what you mean with "distracting passing" and "changing animation", I rewatched the video and can not find any unnecessary or distracting animation BUT yes the PACING IS TOO FAST! I had to slow down to 0.75 (which sounds a bit awkward) and I usually watch videos at 1.25 or even 1.5.
@@AdrianTwarogno it’s perfect. This guys just hatin… lovin the new teams work
thanks
Make it slow
How to say "Azure" - th-cam.com/video/FOXDyRLT6so/w-d-xo.html
Definitely not beginner friendly. You were too fast.
Such a shame that Microsoft who is sponsoring this video keeps Azure OpenAI invitation-only. So they make your video useless to many 👎
How you pronounce azure 😂asua
Its too fast explanation,
azure open ai in a nutshell: how much is 1 + 1? "token limit exceeded try again in i905343895789347895034789050789349078534 seconds"
@Adrian
my scenario is like below :
i have x-train excel file record :
ID Month Year Brand Platform Agency
110 1 2025 pepsi Google Starcom
100 1 2025 pepsi Google Starcom
101 2 2025 pepsi Google Starcom
102 1 2025 CocaCola Facebook Starcom
i have y-train excel file record : (for find)
ID Month Year Brand Platform Agency
102 2 2025 pepsi Google Starcom
i want use azure ai to upload both file and need result accuracy match with percentage or nearby record