For a few year's I have been thinking about making a voice assistant / virtual assistant which you can call and have it answer F.A.Q.'s about a subject. For example a plumbing company would host this voice assistant using a dedicated phone number (through Twilio or similar service) and customers can call to get minor issues resolved. Ideally it would also be able to book appointments and answer billing questions etc.. If you can make a simple version of this it would be super useful. Love your content Tim! been watching you for many years
Hi Tim!!! I´m a Brazilian and you inspired me in the Python's world, I work in the Oil and Gas Industry already for some years and I had already condition to apply part of the knowledge obtained in programming to improve the process and the safety on board of huge FPSOs (Floating Production, Storage and Offloading Vessels which operate far from the coast, in the middle of the ocean, and require a high level of technology). Currenty I´m in Singapore working in huge and new projects for Shell, Exon-Mobil and Equinor, preparing the Ships to operate in Brazil, Mexico and Guiana. It´s incredible how many things we can do when we have the programming fundamental. This changed my life completely!!! You are the best my friend!!! Please, continue inspiring others with your precious information.Thanks for your excellent videos.
Fala Ederson, beleza? Cara, também atuo no ramo de oil e gas aqui no Brasil, tu poderia me passar teu linkedin pra tirar algumas dúvidas de como você evoluiu na carreira sendo que nossa área tem pouca prospecção para tecnologia?
Awesome Tim👏👏 Your voice plays in my head while awake and in my sleep because I follow along on CourseCareers as well. Many thanks for being always clear.💯🤝
Hi Tim, nice video, and nicely explained. Even for those who have never delved into Python, you break in down nicely. I’ve came into python back from when it was still under the radar before AI or LLMs were not on our road map. Another thing that you should include in your video is the network devices that communicate with your ChatBot, ie sensors or network switch box’s etc. Thanks for sharing your video.
Maaaaaan! You saved me from the headache! I was looking for that project's name - I tried a live demo before, and I liked it, but I missed the bookmark, thanks 👍 Subscribed, by the way 😊
YAGV - Thanks Tim, for yet another great video. I always learn a lot. By showing the process (both mental and developmental) that you go through, you convey a lot of subbtle hints, in addition to the primary subject.
best video of all time didnt even explained how you setup your visual studio code top of the line content and than this awkward startig "that you are the best tech youtuber" :D would be nice to know where to paste the first line to install python do i need a wsl terminal or WHICH KIND?!?!??!?!? the promt python -m venv ai dosnt work cuz some people install visual studio code for the first time and didnt know if they have the right terminal to paste the prompt
Nice tutorial ! Thanks. Can you make another video to show how to set conversation flow with voice assistant . ex. assistant should ask questions to user to get required info for some task.
It would be cool to integrate this with the Windows shell so that I can tell it to minimize a window, open a directory, etc. And possibly even more advanced manipulation of the filesystem.
You have talent. Do you think you could make a video on having live word prompting for sales calls? E.g. someone says I’m not interest, goes to vector database to pull from knowledge base for a rebuttal
Mükemmel ve kaliteli bir içerik... kanalınızı yeni keşfettim.manuel olarak altyazı eklediğiniz için otomatik çeviri ile videonuzu izleyebildim. lütfen diger iceriklerinizde de manuel altyazi ekleyiniz. otomatik çeviri ile başka dillerde izlenebilir ve daha geniş kitlelere ulaşabilirsiniz...
Hey Tim i followed your tutorial word for word and my Live-kit playground Room is connected to the Agent, However in my IDE terminal the logs show something entirely different like "from e livekit.agents._exceptions.APIConnectionError: failed to synthesize speech after 4 attempts", "timestamp": "2025-01-21T04:30:53.574064+00:00"}". I think it has something to do with TTS and STT. I'm not sure could you please help me out?
can anyone help me I finished the whole video nad my assistant is not working at all i tried copying his code with my own keys and secrets but it still is not working \
Hey, I really need some help! We’re working on a project with a tight deadline next week, but we keep running into an error saying the OpenAI limit has been reached. It’s really holding us back! If anyone has tips on how to resolve this quickly, it would be a lifesaver. Also, we're looking to add video functions to our project. Any suggestions would be hugely appreciated! Thanks in advance!
Hello we did everything and there are no errors but in the end when we talk it says something about open ai limit reached or something and we can't seem to connect we rlky need this , we have a deadline coming up next week and this is inactive for it , if this works we can make our own version of this and we rely like how it turned out so far , also pls tell us how to add video functions to this, pls reply fast
The limit is reached because OpenAi requires you to pay per use. You only get a free option for the first 3 months of usage of your account, but I suppose you have created your account long ago and that does not work anymore. If you want to fix that you could either add your card information and add like 2e, or something like that, or use Ollama locally. Tim also has a video about using Ollama locally ("AI WebScrapper").
Thanks for the tutorial! I'm making a Corinthian Helmet with a built-in AI of Athena and Ares. Is it possible to use two separate AI personalities in one playground?
Hey Tim, Thank you for this video! Question -> what if i don't want to manually connect and reconnect using the livekit UI? I want something like...i run my python program and it starts running and i just have to speak in my mic locally. Is that possible?
first of all thank you so much you gave me such a wonderful idea and 1 more thing that i just done the whole process but the agent doesn't connecting can you help me why it happening?
Hello Tim, beautiful an useful video as usual, thanks =) I like to ask, what plugins are you using in your visual studio, looks great. Thanks again, regards from Chile.
Did I miss something here, what is it connecti g to to actually change the temperature? Is it some sort of IoT or what? I got confused at some point. Also how do you get it on your phone?
Hey, thank you for the video! I need some help. I'm making an assistant for accessibility, and I want the agent to respond when the user calls it by name (only when the user calls it by name). How can I do this?
nice tutorial ! is there a way to use livekit without their frontend and use only python , for example reading voice with pyaudio and pass that into their stt and tts agents ?
Thank you for your tutorial! I followed the steps, but I encountered an issue when trying to use the TTS-1 model. The error message I received is: livekit.agents._exceptions.APIStatusError: Error code: 403 - {'error': {'message': 'Project `proj_HSXvsRPH4YanhCYP7vPxQmeF` does not have access to model `tts-1`', 'type': 'invalid_request_error', 'param': None, 'code': 'model_not_found'}} I already have a sufficient balance in my OpenAI account. Do you have any suggestions on how to resolve this? Thank you in advance for your help!
hey im having an issue with implementing this. I keep getting the error` api_key is required, or add LIVEKIT_API_KEY in your environment` when i have all of the keys inside my .env file. idk what the issue is as im new to coding
I don't understand there's no difference between using this website and OpenAI's one other than requiring more work, could I run this entirely in python without the need for a website tool? I'm trying to integrate it into a raspberry pi robot.
if i want to know the logic for switching speakers(whether user is speaking or assistent (tts system) is genearting response and to interrupt assistant while user is saying in real time where can i find that (I dont want to use livekit)? Currently i am trying to do it with the help of Silero VAD but would love to know it from you
Hey @TechWithTim I am facing errors while installing livekit-plugins-openai and livekit-plugins-silero (I even tried to install it from requirements.txt) it just shows an error "Getting requirements to build wheel did not run successfully. │ exit code: 1" Please help me if you could!!
hi, same thing in my setup, been dealing with different methods trying to install livekit-plugins-openai wich is the one that triggers the error, did you find a solution?
@@TrinidadIPPC yes the problem is with the python version my python 3.13 was not able to install these dependencies but it solved when I tried with python 3.12. Glad if it helps!!
can we create voice assistant for SQL queries , i gives the input via voice commands and it will be generate the dashboard or visualization for us. or gives the table from the database. can we get videos on this topic? eagerly waiting !
thanks for a great tutorial, been dealing with different methods trying to install livekit-plugins-openai, wich triggers an error about wheel building, any solution? thanks
Great video, I will subscribe and look forward to seeing your other videos. Tried to follow along on this video but ran into a problem in the playground. "agent connected" - does not connect, just spins then ends up False. I will try to figure it out but want to ask this question in your comments section. Would appreciate any help. Thanks,
it is really nice, thanks a lot. How would you do the exact same but with your own UI instaed of the one provided by liveKit? maybe add a part using react, FastAPI?
How do I connect knowledge base I mean RAG to livekits voice agents, In my cases I wanted to add more than 50 pdf files. Could you please provide some information regarding this topic ?
Great Video , Question how can you ad an avatar , so let say i want and to build a Friend assistant hero but i want to talk to the hero in my phone screen . Can you please guide me ? Thank you
Hey I’m super new to coding and I was wondering if I would be able to make it were I can tell a robot arm which way to move and power it on and off with the voice assistant
What kind of voice assistant do you want to make?
J.A.R.V.I.S.
@@UlanKurmanbekov ULTRON?
For a few year's I have been thinking about making a voice assistant / virtual assistant which you can call and have it answer F.A.Q.'s about a subject. For example a plumbing company would host this voice assistant using a dedicated phone number (through Twilio or similar service) and customers can call to get minor issues resolved. Ideally it would also be able to book appointments and answer billing questions etc.. If you can make a simple version of this it would be super useful.
Love your content Tim! been watching you for many years
Can we use local Ollama LLMs to power our AI assistant?
PIP ❌ UV ✅
best developer, best youtuber, best programmer, best explainer.
Hi Tim!!! I´m a Brazilian and you inspired me in the Python's world, I work in the Oil and Gas Industry already for some years and I had already condition to apply part of the knowledge obtained in programming to improve the process and the safety on board of huge FPSOs (Floating Production, Storage and Offloading Vessels which operate far from the coast, in the middle of the ocean, and require a high level of technology). Currenty I´m in Singapore working in huge and new projects for Shell, Exon-Mobil and Equinor, preparing the Ships to operate in Brazil, Mexico and Guiana. It´s incredible how many things we can do when we have the programming fundamental. This changed my life completely!!! You are the best my friend!!! Please, continue inspiring others with your precious information.Thanks for your excellent videos.
Fala Ederson, beleza? Cara, também atuo no ramo de oil e gas aqui no Brasil, tu poderia me passar teu linkedin pra tirar algumas dúvidas de como você evoluiu na carreira sendo que nossa área tem pouca prospecção para tecnologia?
bro you are literally the best developer i have ever seen
Awesome Tim👏👏
Your voice plays in my head while awake and in my sleep because I follow along on CourseCareers as well.
Many thanks for being always clear.💯🤝
Hi Tim, nice video, and nicely explained. Even for those who have never delved into Python, you break in down nicely. I’ve came into python back from when it was still under the radar before AI or LLMs were not on our road map. Another thing that you should include in your video is the network devices that communicate with your ChatBot, ie sensors or network switch box’s etc. Thanks for sharing your video.
Thank You. I am currently investigating a "Python AI voice assistant " for a "speech impaired" friend.
Seriously! Thanks so much!!! I've been wait for a tutorial like this.
This is the best video. Exactly what I was looking for
Subscribed to your channel recently but man am i in love with your content
Maaaaaan! You saved me from the headache!
I was looking for that project's name - I tried a live demo before, and I liked it, but I missed the bookmark, thanks 👍 Subscribed, by the way 😊
YAGV - Thanks Tim, for yet another great video. I always learn a lot. By showing the process (both mental and developmental) that you go through, you convey a lot of subbtle hints, in addition to the primary subject.
best video of all time didnt even explained how you setup your visual studio code top of the line content and than this awkward startig "that you are the best tech youtuber" :D would be nice to know where to paste the first line to install python do i need a wsl terminal or WHICH KIND?!?!??!?!? the promt python -m venv ai dosnt work cuz some people install visual studio code for the first time and didnt know if they have the right terminal to paste the prompt
I’m going to make it and if it works I will like ever video you make for a year cuz I have been looking for this for 2 years
Nice bro imma make jarvis fr
Nice tutorial ! Thanks. Can you make another video to show how to set conversation flow with voice assistant . ex. assistant should ask questions to user to get required info for some task.
Will you do mojo tutorials in the future? Also greate tutorial, your channel truly is one of the best programming youtube channels.
It would be cool to integrate this with the Windows shell so that I can tell it to minimize a window, open a directory, etc. And possibly even more advanced manipulation of the filesystem.
At 17:13, The command will not work, and it gives me this back "[Errno 2] No such file or directory" What can I do
Same problem, did you find a solution?
@@littleicepenguin2370 Ran the code in the terminal for it to work
You have talent. Do you think you could make a video on having live word prompting for sales calls? E.g. someone says I’m not interest, goes to vector database to pull from knowledge base for a rebuttal
Search vaper on Google
Thanks for sharing with us!
I am going to test Portuguese voice.
What is the approach to build such AI assistant in Dart/Flutter? Do I have to run the python code inside Dart by setting up a server?
Mükemmel ve kaliteli bir içerik... kanalınızı yeni keşfettim.manuel olarak altyazı eklediğiniz için otomatik çeviri ile videonuzu izleyebildim. lütfen diger iceriklerinizde de manuel altyazi ekleyiniz. otomatik çeviri ile başka dillerde izlenebilir ve daha geniş kitlelere ulaşabilirsiniz...
Buddy wow i am quick in 19 minutes now i don't know what time i am commenting at but i came at 19 minutes ago.
It was an impressive video!! Thank you so much
at 17:13 im having problems saying my python isnt recognized
Hey Tim i followed your tutorial word for word and my Live-kit playground Room is connected to the Agent, However in my IDE terminal the logs show something entirely different like "from e
livekit.agents._exceptions.APIConnectionError: failed to synthesize speech after 4 attempts", "timestamp": "2025-01-21T04:30:53.574064+00:00"}". I think it has something to do with TTS and STT. I'm not sure could you please help me out?
bro the video is so cool can you suggest me like how do we do this via hugging face or the ollama
can anyone help me I finished the whole video nad my assistant is not working at all
i tried copying his code with my own keys and secrets but it still is not working
\
What kind of error are you getting?
@@TechWithTim I'm getting a Value Error, "ValueError: api_key is required, or add LIVEKIT_API_KEY in your environment"
@@eggsandbrioche you ever fix this? im getting the same error
Hey anyone else get the "ModuleNotFoundError: No module named 'dotenv'" error around 17:33 ?
Hey, I really need some help! We’re working on a project with a tight deadline next week, but we keep running into an error saying the OpenAI limit has been reached. It’s really holding us back! If anyone has tips on how to resolve this quickly, it would be a lifesaver.
Also, we're looking to add video functions to our project. Any suggestions would be hugely appreciated! Thanks in advance!
it's telling me to add the livekit api key to the environment but it's in the .emv file that you instructed. it doesn't see it.
it should be `.env` and not `.emv`
@@kshitijdeshpande6272 I am having this same issue yet is called `.env` any ideas?
Currently building a similar project, Jervis 😊
kindly when you use open api tell us the pipe line in detail like how to use it with free models 🙏
As always great content. Can you provide more content using open source llms please?
Hello we did everything and there are no errors but in the end when we talk it says something about open ai limit reached or something and we can't seem to connect we rlky need this , we have a deadline coming up next week and this is inactive for it , if this works we can make our own version of this and we rely like how it turned out so far , also pls tell us how to add video functions to this, pls reply fast
The limit is reached because OpenAi requires you to pay per use. You only get a free option for the first 3 months of usage of your account, but I suppose you have created your account long ago and that does not work anymore. If you want to fix that you could either add your card information and add like 2e, or something like that, or use Ollama locally. Tim also has a video about using Ollama locally ("AI WebScrapper").
Thanks for the tutorial!
I'm making a Corinthian Helmet with a built-in AI of Athena and Ares. Is it possible to use two separate AI personalities in one playground?
Hey Tim, Thank you for this video!
Question -> what if i don't want to manually connect and reconnect using the livekit UI? I want something like...i run my python program and it starts running and i just have to speak in my mic locally. Is that possible?
first of all thank you so much you gave me such a wonderful idea and 1 more thing that i just done the whole process but the agent doesn't connecting can you help me why it happening?
Same Issue Did you sorted??
@anusha_prathap_world2421 yupe ..you should try again and again it will be connected
Hello Tim, beautiful an useful video as usual, thanks =)
I like to ask, what plugins are you using in your visual studio, looks great.
Thanks again, regards from Chile.
This is so great, but as my startup is for non-english user, can you provide a alternative to the LiveKit?
Did I miss something here, what is it connecti g to to actually change the temperature? Is it some sort of IoT or what? I got confused at some point. Also how do you get it on your phone?
Hey, thank you for the video! I need some help.
I'm making an assistant for accessibility, and I want the agent to respond when the user calls it by name (only when the user calls it by name).
How can I do this?
nice tutorial ! is there a way to use livekit without their frontend and use only python , for example reading voice with pyaudio and pass that into their stt and tts agents ?
Thank you for your tutorial! I followed the steps, but I encountered an issue when trying to use the TTS-1 model. The error message I received is: livekit.agents._exceptions.APIStatusError: Error code: 403 - {'error': {'message': 'Project `proj_HSXvsRPH4YanhCYP7vPxQmeF` does not have access to model `tts-1`', 'type': 'invalid_request_error', 'param': None, 'code': 'model_not_found'}}
I already have a sufficient balance in my OpenAI account. Do you have any suggestions on how to resolve this?
Thank you in advance for your help!
hey im having an issue with implementing this. I keep getting the error` api_key is required, or add LIVEKIT_API_KEY in your environment` when i have all of the keys inside my .env file. idk what the issue is as im new to coding
can we use other opensource models like llama for the same task?
I don't understand there's no difference between using this website and OpenAI's one other than requiring more work, could I run this entirely in python without the need for a website tool? I'm trying to integrate it into a raspberry pi robot.
The agent is not able to understand what I am saying when I try to set it up. I am using the same code as your's.
How would I make this work with self hosted?
Can you make an AI Voice Assistant for a wordpress website, that was feeded with the website and pdfs?
Is there a way livekit agent identifies given wake word and listens only when it is said?
Is there any way to update an old google nest to something like this?
if i want to know the logic for switching speakers(whether user is speaking or assistent (tts system) is genearting response and to interrupt assistant while user is saying in real time where can i find that (I dont want to use livekit)? Currently i am trying to do it with the help of Silero VAD but would love to know it from you
Not sure why I'm getting error, unable to create directory
Hey @TechWithTim I am facing errors while installing livekit-plugins-openai and livekit-plugins-silero (I even tried to install it from requirements.txt) it just shows an error "Getting requirements to build wheel did not run successfully.
│ exit code: 1" Please help me if you could!!
hi, same thing in my setup, been dealing with different methods trying to install livekit-plugins-openai wich is the one that triggers the error, did you find a solution?
@@TrinidadIPPC yes the problem is with the python version my python 3.13 was not able to install these dependencies but it solved when I tried with python 3.12.
Glad if it helps!!
it keeps telling me IndentationError: expected an indented block after function definition on line 11
line 12 initial_ctx
can we create voice assistant for SQL queries , i gives the input via voice commands and it will be generate the dashboard or visualization for us. or gives the table from the database. can we get videos on this topic? eagerly waiting !
Sure you can! You can simply ask the model for a query and then connect it to the database and show the result
@@TechWithTim for this use case ? is any video tutorial available on your you tube channel.
thanks for a great tutorial, been dealing with different methods trying to install livekit-plugins-openai, wich triggers an error about wheel building, any solution? thanks
Can i make this using Gemini ?
Great video, I will subscribe and look forward to seeing your other videos. Tried to follow along on this video but ran into a problem in the playground. "agent connected" - does not connect, just spins then ends up False. I will try to figure it out but want to ask this question in your comments section. Would appreciate any help. Thanks,
Great Video
why is this showing up " from dotenv import load_dotenv
ModuleNotFoundError: No module named 'dotenv'"??
switch from the (base) venv, to the (ai) or whatever venv you named it. in the bottom left in VSCode
can i use llama 3 instead of openai key
Best developer 🫡
hey Tim, can this assignment work on our voice commands. like if we said open Google so can it be,???
Unresponsive playground, what to do? Tried reconnecting, didn't work
Excellent tutorial,thank you, please can we use
Agentic RAG with LiveKit,
wondering how I can connect this locally with ollama?
But why do you cut it when you find out how long it takes to respond in real time?
You can test yourself just don’t wanna make you sit through that in a video
am kinda stuck here not knowing what to do it keeps saying can't find dotenv what should i do pls
did you do load_dotenv()?
for the imports do:
from dotenv import load_dotenv
load_dotenv()
it is really nice, thanks a lot. How would you do the exact same but with your own UI instaed of the one provided by liveKit? maybe add a part using react, FastAPI?
so it can not report anything other than temperature related stuff?
It can do anything you want, I just make a simple example to get you started
how can i build a custom application on that
Bro how do you read my mind? im literally working on the same damn thing using groq and ollama
I don't have money to buy OpenAI key, any other way around?
How do I connect knowledge base I mean RAG to livekits voice agents, In my cases I wanted to add more than 50 pdf files. Could you please provide some information regarding this topic ?
should i have to pay for using openai api?
wait, maybe i can used this for interviews!
hehe
how to make a function that sends the whole conversation to an llm after the conversation ends?
What is this theme you are using?
cant we do this with nodejs?
Nice ❤
Thanks 🔥
@@TechWithTim how to custom voice for this ai?
Voice assistant Can be in french language ?
Yes
How can we capture the audio duration?😊
Bro you are the best in the world ah one time please may you make voice assistant for us please by train your modal and use your audio datasets
I keep getting "fnc_ctx not defined"...
Great Video , Question how can you ad an avatar , so let say i want and to build a Friend assistant hero but i want to talk to the hero in my phone screen . Can you please guide me ? Thank you
Question can I use this to translate
can you make a tutorial on self host mode , it will be cool as well
how do i do this with gemini api key
Can i use it even if am offline
does this support voice interruption?
Hey I’m super new to coding and I was wondering if I would be able to make it were I can tell a robot arm which way to move and power it on and off with the voice assistant
I would use an arduino to do that
How can
you embed this in your website???
does livekit support edge_tts ?
I want to build it with my custom voice, but I don't know how ?
ImportError: cannot import name 'AssistantFnc' from 'api'
Upload a video on how to customize the frontend like video calling app.
How can I use gemini in this project?