This is really cool, I didn’t know this was possible. I already have a few use cases in my business for this. One thing I’d really like to see is a chat GPT chat feature in a power bi report, where Chat GPT could read and understand the data model in power bi and answer questions. I know this feature exists in power BI. What are your thoughts, is it possible and would it be more effective than the built in BI feature?
Interesting when you play around with it but so slow and has it´s limitations only defining data for dim_tables and what not. One question is regarding to get this model to work in the service. Although i have the credentials as anonymous it doesn´t work?
Yes totally agree. Was more of a thought exercise but I do think the concept of applying the capability of the language model in tabular models will get further built out and have practical use cases. Most likely with some sort of native Microsoft connector not through the straight API. I have another video I’m working on that will cover use in service, just need to find the time to get it out!
Can we give an option to the end users, asking questions to chat GPT as an end user after we publish Power BI Report to the services, this will be very useful. ( Just like Q & A , Chat GPT Interface should be created )
Not that I’ve figured out yet. Closest thing in general use is the powerbi Q&A functionality but I think that copilot is being released in the next 30 days to general public. Should do exactly what you are asking for. I’ll be sure to check it out and make a video on it
@@Ptarmy_ from IPython.display import Markdown, display import os import pandas as pd from llama_index.experimental.query_engine import PandasQueryEngine os.environ["OPENAI_API_KEY"] = "XXX" # Test on some sample data df = pd.DataFrame( { "city": ["Toronto", "Tokyo", "Berlin"], "population": [2930000, 13960000, 3645000], } ) query_engine = PandasQueryEngine(df=df, verbose=True) response = query_engine.query( "Summarize the data and give insights", ) print("Response is like this: {}".format(response)) PowerQuery can accept dataset and we can feed question but i cant get this working in powerquery
I am able to create a custom funtion.But when i am giving an input and press the invoke button .. I am getting an credentials error with the data source I tried many ways but every time getting the same error. Please help..
This is helpfull .Thanks a lot. Can we analyze are dataset by the help of chat gpt in power bi. Like I want to know my top 10 customer in the power BI report by chatgpt.How can we implement this.I guess this is possible we can easily do this in SAC.Could you please suggest?
thanks so much man... because of you i'm going be the guy of my office today
Awesome, great video! Thanks for sharing!
Hey @Ptarmy, Thanks for this video. Wonder how do you pass an entire dataset and a question
Where did you get the code you posted at 1:29?
platform.openai.com/docs/guides/gpt
This is really cool, I didn’t know this was possible. I already have a few use cases in my business for this.
One thing I’d really like to see is a chat GPT chat feature in a power bi report, where Chat GPT could read and understand the data model in power bi and answer questions. I know this feature exists in power BI.
What are your thoughts, is it possible and would it be more effective than the built in BI feature?
If you just paste your api key there is it secure from the readers of a published report?
Yeah from powerbi service I don’t see how a reader can get into the power query side and see your APIkey
Interesting when you play around with it but so slow and has it´s limitations only defining data for dim_tables and what not. One question is regarding to get this model to work in the service. Although i have the credentials as anonymous it doesn´t work?
Yes totally agree. Was more of a thought exercise but I do think the concept of applying the capability of the language model in tabular models will get further built out and have practical use cases. Most likely with some sort of native Microsoft connector not through the straight API. I have another video I’m working on that will cover use in service, just need to find the time to get it out!
Can we give an option to the end users, asking questions to chat GPT as an end user after we publish Power BI Report to the services, this will be very useful. ( Just like Q & A , Chat GPT Interface should be created )
Hi. This is really cool. Can you also generate the visualisations based on your data using chat gpt?
No direct interaction with the canvas and creating visuals I have found yet!
Can we use chatgpt to retrieve answers or reports from a uploaded table in powerbi
Not that I’ve figured out yet. Closest thing in general use is the powerbi Q&A functionality but I think that copilot is being released in the next 30 days to general public. Should do exactly what you are asking for. I’ll be sure to check it out and make a video on it
@@Ptarmy_
from IPython.display import Markdown, display
import os
import pandas as pd
from llama_index.experimental.query_engine import PandasQueryEngine
os.environ["OPENAI_API_KEY"] = "XXX"
# Test on some sample data
df = pd.DataFrame(
{
"city": ["Toronto", "Tokyo", "Berlin"],
"population": [2930000, 13960000, 3645000],
}
)
query_engine = PandasQueryEngine(df=df, verbose=True)
response = query_engine.query(
"Summarize the data and give insights",
)
print("Response is like this: {}".format(response))
PowerQuery can accept dataset and we can feed question but i cant get this working in powerquery
I am able to create a custom funtion.But when i am giving an input and press the invoke button .. I am getting an credentials error with the data source I tried many ways but every time getting the same error.
Please help..
Re create a new api key and try again. Also watch the openai uptime on their api. It was down yesterday for a short while
This is helpfull .Thanks a lot.
Can we analyze are dataset by the help of chat gpt in power bi.
Like I want to know my top 10 customer in the power BI report by chatgpt.How can we implement this.I guess this is possible we can easily do this in SAC.Could you please suggest?
Could you pleaee share the python code?
All code and templates available for free at ptarmy.com/joinus