A detailed video about how to connect Flowise to a (local) PgAdmin database would be super valuable. I haven't come across a clear tutorial on this topic yet, so I think it will benefit many!
I've found your flowise series extremely helpful thank you! Would love more of these types of videos including records manager how to. I'm keen to learn all of the different features of flowise :)
Great video! And yes please! More Record Manager Videos. This has always been my issue. Almost every week I need to update a very specific information of my data set. I only need to insert the updates, not everything again.
Great video! WOuld be nice to see more on record manager and how to feed docs from a cloud drive directory, so when the directory is updated, your bot will be updated.
Yes i would like to see that video on how to manage records! also it would be great another one on how to use the endpoits provided by flowise to pass audio and image files :)
Great session thx Leon, would really be interested how to prevent duplication into Pinecone when you upsert, currently I have to delete all Pinecone records. Would also like it to know how a user can use multiple Selectors targeting different data sources. Thanks once again for your informative session greatly appreciate the effort going into creating this material, I love all your Flowise training material 👍👍would love to see more use cases in the future.
Thanks for all the videos...I would really appreaciate a video explaining the memory manager. I am also wondering how happy a Pinecone database is to just have maor data added to it via subsequent upserting of documents
Great video! 💯 Idea for the next video - would be great to know how we can set up a chatbot where our client can manage his RAG documents by its own without going to the main flowise chatbot interface. Is there a way to just have some small backend page for the client or similar? Also can we set up chatbot to rescan to RAG some sources like webpage at intervals. Thanks!
Great suggestion! It is very possible. I have a video on using the Flowise API's for executing upserts, but deleting and managing knowledge bases are definitely managed outside of Flowise (at this stage). I think this is a great project idea. I'll work on a full tutorial.
As usual .. very nice and useful .. Flowise is full of secrets and it is one of the best low-code tools.. Thank you... It seems I have became addicted to this wonderful channel ♥
A cool tutorial I would like to see is the record manager and also something about how to use databases to use past conversations as context on new conversations.
Hello Leon, thank you for your content! I really need a video explaining how to read csv files such as a Sales file, where you have different columns with different data types, and you can ask the chatbot for information and for certain operations using the data from the csv.
I would like more information about RAG, Record Management, using mySQL on a domain server, and end-to-end encryption from the user to the encryption of the database. By the way, this is an excellent video.
Great video! I really enjoy all your videos because you know how to explain everything very well. I know that your channel will grow a lot in a short time! My only question, is it enabled in the Flowise API the possibility to pass images and audios through the endpoints?
Awesome Video and super useful. Keep more vidoes like these coming. Openai has added streaming to assistants now so is it possible to do the same inside of Flowise?
Yes please, a full video for SQL Record manager and see how it acts would be helpful. Also would like to ask you if you could talk about the fallbacks. As you may already know, Copilot studio (old Power Platform) once it has a fallback you can manage it with topics or entities. Bsed on the reply or comment of the user. In this example that you are showing you can "only" take a look of what the user is answering and based on that, what do you do? is there any way to take actions to the questions the user makes? Should we manage this with a prompt engineering before or after gathering information? as far as I Know, f,owise or langchain do not reflect tree decisions. Also a video of a mix BOT would be amazing. Like if a user makes a question, the BOT answers with Vector Database or making an SQL query if the info is on a Database, would be that possible? Thank you for your great job, Leon.
Your right, Flowise is not a "tree decision" based platform - you might want to consider something like Voiceflow or Botpress for guiding the user down different paths based on their intent / messages. The closest you'll get in Flowise is function calling / tool use using Agents.
This is awesome 😊 May I ask that which version of flowise are used in this video because I did not find feedback or record manager function in previous version.
How would you rank the following tools: Flowise, Vectorshift, Stack AI, Buildship and Relevance AI? I'm still deciding which one to pick as the go to option for my projects.
Thank you very much for your valuable resources. They helped a lot. Recently, I'm trying to pass Audio inputs to a chatflow (with one Open AI assitant) that has voice input enabled (Whisper). However, my AI assistant only accepts text inputs. I use python and followed the documentation, but no luck so far. Can you please advice on me this ? If assistant nodes can't accept audio inputs, what is the method to do this via python ? Thank you in advance !
You're welcome 🤗. I'm not sure what you mean since a cloud application will not be able to access a local model running on your machine. If you're referring to Open source models then you might want to check out my Huggingface video 👍
Just had a go with MYSQL. Had to enter credentials for MYSQL and a tablename. The upsert automatically created the table. I uploaded a large text file HS1.txt (fullmode) and I got the upsert records added 5677. I then replaced the HS1.txt with HS2.txt and selected Incremental. This overwrites the vector database ( I see the vector count go back to 0 then updating). What am I missing here. The issue I was having is that if I try to upsert lots of large txt files (56 meg in total) the upsert would time out.
@leon amazing brother thanks. Do you know if there is a way to create a chatbot which takes in documents -> embeds -> and has that as context? say for example for a hiring chatbot that could take in your CV and connect you with a job offering? Thanks dude!
Hi Leon, can we attach an external persistent database to save the chat flows on huggingface. The reason for asking was to also check on how can we update flowise on hugginface - I installed using the docker script initially, but wanted to maintain the chatflow in a persistent storage.
thanks for the video, do we have the ability to upload other files besides images? For example if the user wants to upload a csv or pdf to analyze data that is not needed to be stored in vector database. This would be like what we can do in chatgpt4o
How to add "Upload file" to the chat window? For example, I want to ask the user to upload a pdf and answer his questions based on the content of the pdf document.
did you found the way to create a flow with image recnonigtion and API? for example read a pdf and send some info to googlesheets or other app via POST API?
I don't understand how Record Managers work, or better, I think I do, but that would be a way too simplistic approach. Assume that I have a pdf and add a paragraph somewhere in it. This newly added paragraph, depending on the text splitter and chunk_size/overlap, might have a waterfall effect of grouping the following (unchanged) paragraphs in a different way, and therefore generate completely different hashes. What can Record Manager do to handle this use case and provide a proper ingestion mechanism? None of the offered features seem production ready enough to handle this very simple and common use case.
Great video as always. I would like to add local whisper AI to my flow. Any way to do this, instead of pinging their API. Would like to keep my files and transcriptions on my machine. Thanks.
Thanks Leon this was really helpful. For the rate limit, I have tried even before now it didn't work. From fkowise documentation I think this can be done from the cloud hosting provider like Render or Railway. Please can you do a video on how to setup this on Render and Railway
is there a video coming about this, this could be a monster if you know how to built it with GoHighLevel If you want, I can give you a snapshot of how I already did it and even explain how it was done! I wanna help you in any way I can '
I have a flowise installation based on render. When I click on the "Settings wheel" on the right above I only get an "about flowise link" Not settings at all. Did flowise change this and delete this Speach to Text function? Or where can I find it?
@@leonvanzyl sorry, I just found the solution. The settings are not shown as long as I did not open a chatflow, so they are not able to set in general. Thanks
Hi! I´m having issues on how to set up Pinecone in Flowise, couldn´t find anymore the information of the Environment information on Pinecone to inser in the flowise upsert, does anyone know how to deal with that ?
@@leonvanzyl thank you so much for the reply! your videos are excelent! about the issue i mentioned before, I´m still having it, as i do not have the environment named on pinecone, and flowise requires one at the credentials os pinecone, i´m receiving the following answer at the flowise chat: "PineconeArgumentError: The client configuration must have required property: environment." what could i do to solve it ?
@@leonvanzyl thx for the response. It is more that Render is only secured by one type of password user combination. If you guess the password you have excess to all infrastructuur. Also, is it possible to take SharePoint as a docs loader,?
Thanks! Amazing video as always. Do you know if there's a way to include persistent buttons on the chat? I really like how Brainglue is doing it (ua-cam.com/video/fvWvg8jTAVs/v-deo.html) when you can pick different brains or choose to save a reply for later.
A detailed video about how to connect Flowise to a (local) PgAdmin database would be super valuable. I haven't come across a clear tutorial on this topic yet, so I think it will benefit many!
Great suggestion!
Yes record manager please ! And more fun with this Quick tips format ! Thanks for all your precious content )
Will do
Seconded the request for a record manager
@@leonvanzyl +1 on record manager tutorial :)
Please do a video on ReAct Agents/Reason Action prompting in Flowise!
Yes, Leon. This type of video is extremely helpful. Thank you for all that you do, especially your work with Flowise.
I've found your flowise series extremely helpful thank you! Would love more of these types of videos including records manager how to. I'm keen to learn all of the different features of flowise :)
Hi very good video ! Looking forward to the record manager video set up
Great video! And yes please! More Record Manager Videos. This has always been my issue. Almost every week I need to update a very specific information of my data set. I only need to insert the updates, not everything again.
Great video! WOuld be nice to see more on record manager and how to feed docs from a cloud drive directory, so when the directory is updated, your bot will be updated.
Yes i would like to see that video on how to manage records! also it would be great another one on how to use the endpoits provided by flowise to pass audio and image files :)
Yes please to the Video on the record Manager! Thank you soooo much for sharing this!🎉❤
Thanks !! more Flowise, text to speech please!
Great session thx Leon, would really be interested how to prevent duplication into Pinecone when you upsert, currently I have to delete all Pinecone records. Would also like it to know how a user can use multiple Selectors targeting different data sources. Thanks once again for your informative session greatly appreciate the effort going into creating this material, I love all your Flowise training material 👍👍would love to see more use cases in the future.
Yup.. more of these please. Thank you.
Will do!
this is great updates! I like the Record Manager one!
Thanks for all the videos...I would really appreaciate a video explaining the memory manager. I am also wondering how happy a Pinecone database is to just have maor data added to it via subsequent upserting of documents
Great video! 💯 Idea for the next video - would be great to know how we can set up a chatbot where our client can manage his RAG documents by its own without going to the main flowise chatbot interface. Is there a way to just have some small backend page for the client or similar? Also can we set up chatbot to rescan to RAG some sources like webpage at intervals. Thanks!
Great suggestion! It is very possible. I have a video on using the Flowise API's for executing upserts, but deleting and managing knowledge bases are definitely managed outside of Flowise (at this stage). I think this is a great project idea. I'll work on a full tutorial.
I would love a video on record managers.
👏👏👏 record manager PLEASE…this is cool…
This should be your #1 video man! These features make a lot of difference in production and for good user experience
As usual .. very nice and useful .. Flowise is full of secrets and it is one of the best low-code tools.. Thank you... It seems I have became addicted to this wonderful channel ♥
Thank you!
Great video and great format. An additional one on setting up the record manager DB's would be helpful. Thank you
Will do
Yes. I like this kind of quick tip video!
Great video Leon, thnx!
Glad you enjoyed it
Love your vids! Short, efficient, initiated. Thanks!
You're welcome 🤗
Yes, more videos like this one! Thanks for all of the great info on Flowise!
A cool tutorial I would like to see is the record manager and also something about how to use databases to use past conversations as context on new conversations.
Thank you for the feedback.
For long term memory, stored in a Redis database, check out this video.
ua-cam.com/video/xJ3iohxHIGo/v-deo.html
Hello Leon, thank you for your content! I really need a video explaining how to read csv files such as a Sales file, where you have different columns with different data types, and you can ask the chatbot for information and for certain operations using the data from the csv.
You're welcome.
Simply use the CSV document loader. Because CSV is structured data, loading the data and chatting to it is really effective.
Great video Leon, immensely helpful
Glad it helped
Great video! Would love to see Record Manager video!
Working on it 😄
I would like more information about RAG, Record Management, using mySQL on a domain server, and end-to-end encryption from the user to the encryption of the database.
By the way, this is an excellent video.
great video as always
I appreciate that
I highly appreciate your time in creating such high-quality videos
Great video! I really enjoy all your videos because you know how to explain everything very well. I know that your channel will grow a lot in a short time!
My only question, is it enabled in the Flowise API the possibility to pass images and audios through the endpoints?
Thank you for the kind words.
You can definitely pass audio and images via API.
Your flowise series is extremely helpful thank you! Would love more of these types of videos
You got it!
Great less known features, tips and tricks. Well presented.
Great as always. Yes to your questions. Please do a video about record manager.
Leon The flowise community has been asking a lot for a tutorial using flowise's streaming socket. This would be fundamental and extremely necessary.
Noted.. sounds like a good tutorial opportunity
Your videos are very interesting, thank you for the posts, they help me a lot to improve my chatbot.
Your videos are top-notch sir. Yes! On the detailed video for the record manager.
Thanks , Sir.
Most welcome
Thanks, Leon. I'm learning a lot from your tutorials. Speaking about features, do you plan to make content about Agentflows? They're in beta now.
Absolutely! The video should be out in the next few days.
Wow never knew all of this, I just use the APIs! This is cool
Yes, very appreciated, thanks 😊
great! yes, you can do this more often! great content!
Thanks! Will do!
Awesome Video and super useful. Keep more vidoes like these coming. Openai has added streaming to assistants now so is it possible to do the same inside of Flowise?
I actually haven't tested streaming for OpenAI Assistants yet.. thanks for reminding me to check it out :-) .
I totally didn't know about Speech-to-text feature in Flowise... thank you
You're welcome 🤗
Super Awesome Video. Please make a video on Record Manager.
Will do
yes i would more flowise video please I am learning a lot thank you for that
Will do
Leon you are certainly the very best! Thank you!
yesss I like to see more contents like this
Great Video as always! Well done!
Yes!!! Video about record manager and maybe about VectorAdmin to see how to manage all the records, pinecone is a pain in the bum bum lol
Nice tips!!!!
Yes please, a full video for SQL Record manager and see how it acts would be helpful.
Also would like to ask you if you could talk about the fallbacks.
As you may already know, Copilot studio (old Power Platform) once it has a fallback you can manage it with topics or entities. Bsed on the reply or comment of the user.
In this example that you are showing you can "only" take a look of what the user is answering and based on that, what do you do?
is there any way to take actions to the questions the user makes? Should we manage this with a prompt engineering before or after gathering information?
as far as I Know, f,owise or langchain do not reflect tree decisions.
Also a video of a mix BOT would be amazing. Like if a user makes a question, the BOT answers with Vector Database or making an SQL query if the info is on a Database, would be that possible?
Thank you for your great job, Leon.
Your right, Flowise is not a "tree decision" based platform - you might want to consider something like Voiceflow or Botpress for guiding the user down different paths based on their intent / messages.
The closest you'll get in Flowise is function calling / tool use using Agents.
@@leonvanzyl thank you for your response
This is awesome 😊
May I ask that which version of flowise are used in this video because I did not find feedback or record manager function in previous version.
This is the latest version.
How would you rank the following tools: Flowise, Vectorshift, Stack AI, Buildship and Relevance AI? I'm still deciding which one to pick as the go to option for my projects.
Thank you very much for your valuable resources. They helped a lot. Recently, I'm trying to pass Audio inputs to a chatflow (with one Open AI assitant) that has voice input enabled (Whisper). However, my AI assistant only accepts text inputs. I use python and followed the documentation, but no luck so far.
Can you please advice on me this ? If assistant nodes can't accept audio inputs, what is the method to do this via python ?
Thank you in advance !
Yes please a video about record mansger and images generation..
Hey Leon… when did you get that fancy UI Chatbot :)
Leon, thank you this is a great video. Please tell me how to connect to the local Ai model from the Flowise cloud application.)
You're welcome 🤗.
I'm not sure what you mean since a cloud application will not be able to access a local model running on your machine. If you're referring to Open source models then you might want to check out my Huggingface video 👍
Is it possible to do text-to-speech in Flowise? Let's say that what the chatbot says is read out loud? Could you make a video about that?
please, a video about Record Manager
Excellent tuto! Would you consider doing some on baby agi and agent gpt? Do you have any insight on when crew ai might make it's way into flowise?
Great suggestions. I'll see what I can do 👍
Would love to see the record manager using a MYSQL database. Been struggling with updating pinecone for a while.
Just had a go with MYSQL. Had to enter credentials for MYSQL and a tablename. The upsert automatically created the table. I uploaded a large text file HS1.txt (fullmode) and I got the upsert records added 5677. I then replaced the HS1.txt with HS2.txt and selected Incremental. This overwrites the vector database ( I see the vector count go back to 0 then updating). What am I missing here. The issue I was having is that if I try to upsert lots of large txt files (56 meg in total) the upsert would time out.
Please do make a video on record manager... Thanks
@leon amazing brother thanks. Do you know if there is a way to create a chatbot which takes in documents -> embeds -> and has that as context? say for example for a hiring chatbot that could take in your CV and connect you with a job offering? Thanks dude!
Hi Leon, can we attach an external persistent database to save the chat flows on huggingface. The reason for asking was to also check on how can we update flowise on hugginface - I installed using the docker script initially, but wanted to maintain the chatflow in a persistent storage.
so fresh a video
Thank you
Loving your work! Is it possible to do a document management on MongoDB?
Not at this stage. Seems to be SQL databases only.
thanks for the video, do we have the ability to upload other files besides images? For example if the user wants to upload a csv or pdf to analyze data that is not needed to be stored in vector database. This would be like what we can do in chatgpt4o
How to add "Upload file" to the chat window? For example, I want to ask the user to upload a pdf and answer his questions based on the content of the pdf document.
did you found the way to create a flow with image recnonigtion and API? for example read a pdf and send some info to googlesheets or other app via POST API?
I want to customize these features myself. How can I change the title and welcome text in the chat feedback?
I don't understand how Record Managers work, or better, I think I do, but that would be a way too simplistic approach. Assume that I have a pdf and add a paragraph somewhere in it. This newly added paragraph, depending on the text splitter and chunk_size/overlap, might have a waterfall effect of grouping the following (unchanged) paragraphs in a different way, and therefore generate completely different hashes. What can Record Manager do to handle this use case and provide a proper ingestion mechanism? None of the offered features seem production ready enough to handle this very simple and common use case.
Great video as always. I would like to add local whisper AI to my flow. Any way to do this, instead of pinging their API. Would like to keep my files and transcriptions on my machine. Thanks.
please do video about record managers!
Already got you covered 😁.
ua-cam.com/video/sNk6-ISi7i4/v-deo.html
Thanks Leon this was really helpful. For the rate limit, I have tried even before now it didn't work. From fkowise documentation I think this can be done from the cloud hosting provider like Render or Railway. Please can you do a video on how to setup this on Render and Railway
Could be a bug in the newer releases actually. I'll make the Flowise team aware.
@@leonvanzyl Thank you. I shall follow-up with you to know thier response
Record manager
Record manager view please. 🙏
How do you enable these things like feedback and speech to text in Flowise embed?
Video about how to use a chain of agents
1 Mega Agent and some mini agents
is there a video coming about this, this could be a monster if you know how to built it with GoHighLevel
If you want, I can give you a snapshot of how I already did it and even explain how it was done! I wanna help you in any way I can '
can you make a video regarding redis cache memory
Like this one? 😊
ua-cam.com/video/xJ3iohxHIGo/v-deo.htmlsi=zL0pgr6knJagBSf_
Are we able to create of embed forms or carousels? under FlowiseAI?
Not at this stage.
I think you're looking for something like Voiceflow or Botpress.
Please let me know if it possible when Website is load and Flowise ai give us a Proactive Message?
please respond me as soon as possible
I have a flowise installation based on render. When I click on the "Settings wheel" on the right above I only get an "about flowise link" Not settings at all. Did flowise change this and delete this Speach to Text function? Or where can I find it?
Which version of Flowise are you on?
They did not remove the settings, by the way. Just wondering if you need to upgrade.
sorry I found the solution;-) All the settings are only available within an open chatflow, not on the general page.
@@leonvanzyl sorry, I just found the solution. The settings are not shown as long as I did not open a chatflow, so they are not able to set in general. Thanks
Hi! I´m having issues on how to set up Pinecone in Flowise, couldn´t find anymore the information of the Environment information on Pinecone to inser in the flowise upsert, does anyone know how to deal with that ?
I have a dedicated Flowise RAG video which includes Pinecone and the Upsert feature.
ua-cam.com/video/V7uBy3VQJAc/v-deo.html
@@leonvanzyl thank you so much for the reply! your videos are excelent! about the issue i mentioned before, I´m still having it, as i do not have the environment named on pinecone, and flowise requires one at the credentials os pinecone, i´m receiving the following answer at the flowise chat: "PineconeArgumentError: The client configuration must have required property: environment."
what could i do to solve it ?
Hi Leon, is there a way to host Flowise on Azure ? Think Render is not secure enough to be honest.
Yes, there are instructions in the Flowise docs.
I haven't personally had issues with my instance on Render which has been running for 8 months 😅.
@@leonvanzyl thx for the response. It is more that Render is only secured by one type of password user combination. If you guess the password you have excess to all infrastructuur. Also, is it possible to take SharePoint as a docs loader,?
Hi Leon, do I need a paid Anthropic account to enable the image upload function?
The free credits work too
It is not showing in my flowise?
Hi Leon, Cannot message you VIA X but sent you one via your site looking forward to your response..
Let me check
cant get acces to flowise
What you do you mean?
@@leonvanzyl waiting list or smthing
Thanks! Amazing video as always.
Do you know if there's a way to include persistent buttons on the chat?
I really like how Brainglue is doing it (ua-cam.com/video/fvWvg8jTAVs/v-deo.html) when you can pick different brains or choose to save a reply for later.
You're welcome. Not aware of such a feature at this stage.
This is gold🎉
Glad you liked the video