Відео

thanks so much! Appreciate the kind words -- VoiceFlow is primarily designed for chat rather than voice, contrary to what their name suggests. While they're beginning to enter this space, I would say the list of viable alternatives we presented would be: Retell Thoughtly

@@Mark_Kashef Many thanks Mark for the prompt response!

thanks so much! appreciate the kind words

Thank you so much for the kind words! Appreciate the feedback 🦾

appreciate you king!

Thanks so much Eamonn! Glad you found it valuable 🦾 much appreciate the feedback

hey! Great catch yes, I totally forgot to include that. it's not 100% reliable at all, but if you go on your GPT > Privacy Settings then select Always Allow, it should 'help' Screenshot: paste.pics/SANDR

I really hope so haha! Thanks for the positive vibes.

I believe they have AWS data centres in the US and Europe -- would be worth confirming in the discord!

hey! This use case primarily for productivity based workflows within an organization -- instead of setting up agents using frameworks that are hard to configure or monitor for performance like Relevance AI or Swarms etc, building these tools act mini assistants with oversight. For your voice example, you could build a GPT that kicks off a webhook to have an internal support agent call the user, or you could also manually initiate calls to a list you've predefined in the Custom GPT -- imagine we had Phone GPT we setup, and we provide it with names and numbers we want it to call; this could link to a Make automation that iterates through the list you've made, and send calls to each one. Big picture, you can kickoff automations and workflows from these GPTs when used in tandem. Similar to piranhas, a single custom GPT is useful but not a full solution, but a group of them can be 🐟

If the GPTs are not made fully accessible and build by another party, then you’re correct - a workaround is creating a copy of said GPTs so that you’re able to access them

@@Mark_Kashef thanks. I didn´t full understand your response but I will continue learning the best way. Good work.

Thanks so much! Appreciate the comment 🦾

So in the video I mentioned that if you go to the GPT Store, you’ll find a GPT made by ChatGPT called ActionsGPT you can use for free - this is an awesome first step to making a schema. You might need to go back and fourth a few times

thanks so much for the kind words! OpenAI should be sponsoring me for that one hahah, appreciate you sharing 🦾

my absolute pleasure -- thanks so much for the kind words!

my pleasure! you pushed me to figure it out over my morning cereal. I made the silence duration a bit longer so you can tweak that as you wish -- it should greet you first and not be able to get interrupted until it's done its opening message: bit.ly/3Uzpmai I'm travelling to Qatar in a few weeks so the prompt is going to be a travel guide haha -- used it myself for a bit. Let me know if it helps!

@@Mark_Kashef Great solution! thanks a lot!!! Have a nice trip!

@@cowfactory4903 You got it! :) and thanks!

Yes indeed! You can literally hook up 5-6 actions to a single GPT, then create 5 of them 🦾🧐 Thanks for the positive encouragement as usual Nataliya ✌🏽

Can I be the CEO? 🤣

Haha thank you! It’s all good, I’m thankful for the audience I have who appreciate quality 🦾

I use something called Screen Studio!

it's not reliable at all, but if you go on your GPT > Privacy Settings then select Always Allow, it should 'help' -- not 100% reliable Screenshot: paste.pics/SANDR

thank you so much! Pleasure to have your sub, appreciate it

awesome feedback; did it again in my most recent video so I'm sure you're not the only one :) appreciate the comment

thank you so so much! happy to hear it was clear and delivered for you 🦾

Love reading comments like these - appreciate your words!

interesting recommendation -- have never looked into langflow; can you connect to any service even if it isn't in their integration repo?

@@Mark_Kashef ​ I've had a look at their doc and can see that you can connect to what it seems any api It also seems easier to connect to an army of bots together from seeing other videos. It also has premade integrations vector databases including pinecone, Embeddings models. Also check out AI/ML AI, you pay once for multiple models.

I've had a look at their doc and can see that you can connect to what it seems any api It also seems easier to connect to an army of bots together from seeing other videos. It also has premade integrations vector databases including pinecone, Embeddings models. Also check out AI/ML AI, you pay once for multiple models.

That’s how I started off if you watch my original meta promoting videos - waiting for text to video to come out and then I have something pretty special coming

My pleasure!! Super pumped it’s helpful my friend 🦾

thank you SO much! I appreciate you saying that, always trying to push the boundaries 🦾

@@Mark_Kashef keep doing what you’re doing🙌🏼

hahahah thanks so much for the kind words Tony -- I know I seemingly bounce between different topics, but I'm usually doing that while prepping for stuff like this!

it's all about the devil in the details my friend!

Pleasure to e-meet you! Appreciate your amazing feedback and pumped to see what else you guys have in store 🦾

hey! Honestly loading 100 scripts might do more harm than good; it can introduce a lot of noise, and it won't necessarily help the LLM learn since it'll be scanning over it briefly. You're better off providing these scripts in batches to an LLM and asking it to keep track of style, language, registry, and structure nuances so it can create a mini summary of its understanding of those scripts that's truly generalizable versus just throwing in the scripts and expecting it to capture all those micro nuances with a quick multi second scan

@Mark_Kashef do u have a solution ill be grateful 🙏

​@Mark_Kashef do u have a solution I'll be grateful 🙏

thanks non-techy King 👑

I'm glad this was helpful! In terms of context windows, they're the same as before, no changes with Spaces unfortunately from what I can see -- one little hack I found is that if you avoid uploading files and you paste chunks of text directly in the chat instead, you can artificially extend this. But obviously this isn't fullproof -- yet another reason to keep using GPTs haha.

@@Mark_Kashef Thanks for that! I'll definitely try it.

Ever since this video dropped, it indeed became a paid feature unfortunately

@@Mark_Kashef well that unfortunate indeed

hey Jim! Happy to do a paid consult, but before that, did you have a chance to download the Make file from my Gumroad? It’s pre loaded with the prompts and will automatically populate when you import the blueprint (json file in my link). I also recently updated exactly how to update the JSON step at the end of

Hey Jonathan! I made it available for free through the Gumroad link in the description; the Make blueprint you can import is there, and it comes with the preloaded prompts 🦾

@@Mark_Kashef Thanks a lot for sharing the link and making it available for free - that's super generous! I’ll dive into the blueprint and prompts and see how they work. Looking forward to experimenting with it and hopefully adding some exciting new services to my agency. Appreciate the support!

In the logs, below, it'll usually tell you what the underlying issue is; usually; API keys are mis-entered, or perhaps additional characters have been added to the code that are clashing its deployment. If not, I've had a few times where Replit itself was backed up, so I waited an hour or two and the deployment worked with no changes

Thank you Jerry! 🦾

thank you sir! Congrats on your first video 😉

yes! You just have to specify in the underlying prompt;

@@Mark_Kashef thank you!

Your use case is precisely why one of our primary clientele for voice solutions are in the healthcare sphere 🩺 That being said, public sector typically lags between 3-5 years for most bleeding edge tech. Voice as a modality is exceptionally better but still not perfect, so perhaps there’s some apprehension until it matures.

Thanks for the feedback! I appreciate you sharing; noted for next time 🦾

Thanks! Daily is pretty rough as there’s a lot of pre work I do for the videos but I’m going to try and stick to the 2 a week 🦾 On your second request, I’m likely going to record a generative AI for beginners near mid to end of November leading into a learning roadmap for 2025

Thanks for your comment! In terms of the requirement from the client, you'll have to convert the CSVs into PDFs to be ingested by Perplexity properly. In terms of the volume of files, I highly recommend you watch my '100% accurate answers' video as I think using Pinecone Assistant might help quite a bit as long as you can convert said Excel files into PDFs in a way where you don't lose the dimensionality and meaning of columns etc.

If you want to do the easy way, as Mark said, Pinecone assistants are a great choice, but I you want something very customized then build a RAG system (using Langchain or LlamaIndex or something similar), that'll get you the best possible results.