Search for a command to run...
Principal Program Manager, Microsoft Fabric CAT helping users and organizations build scalable, insightful, secure solutions. Blogs, opinions are my own and do not represent my employer.
No comments yet. Be the first to comment.
Last week Microsoft released an open-source text embedding model called Harrier in three sizes- 270M, 0.6B and 27B. I have been testing it in my RAG pipeline and so far it has crushed all my metrics.
For Power BI Copilot and Data agents with semantic models, you must use Prep Data for AI configuration to ground the responses in the context added in Prep for AI. In this blog, I will show you how yo
At FabCon Atlanta last week, the updated notebook Copilot for data engineering and data science was announced. It brings agentic capabilities to the Copilot and is much more intelligent and Fabric-awa
Using Fabric data agent Python SDK
Use workspace monitoring to monitor MCP server use
This is a second blog in a row inspired by Edgar Cotte (Sr PM, Fabric CAT). At a recent RTI workshop, he shared a handy KQL function geo_info_from_ip_address which retrieves geolocation based on IP address. You can read more about it here.
A few months ago I wrote a blog on getting workspace activities using Semantic Link Labs. I have been using it on one of my workspaces in personal tenant to generate activity data. I retrieve the logs for each day and save it to a lakehouse. The activity event logs have Client IP address and I wanted to try above function on this field. So I shortcutted the delta table to a KQL table in an Eventhouse and it worked like a charm. Super helpful for auditing & monitoring the workspace activities.
## GET ACTIVITIES FOR THE LAST 7 DAYS AND SAVE IT TO A LAKEHOUSE USING POLARS
## USING PYTHON NOTEBOOK BELOW
from datetime import datetime, timedelta
import pandas as pd
import sempy_labs as labs
import polars as pl
#last 7 days
N=7
activities = []
for n in range(N):
day = datetime.now() - timedelta(days=n)
start_of_day = day.replace(hour=0, minute=0, second=0, microsecond=0).strftime('%Y-%m-%dT%H:%M:%S')
end_of_day = day.replace(hour=23, minute=59, second=59, microsecond=999999).strftime('%Y-%m-%dT%H:%M:%S')
df = labs.admin.list_activity_events(
start_time=start_of_day,
end_time=end_of_day
)
activities.append(df)
final_df = pd.concat(activities)
pl_df = pl.from_pandas(final_df)
## change to your abfss
pl_df.write_delta("abfss://Sales@onelake.dfs.fabric.microsoft.com/MyLakehouse.Lakehouse/Tables/dbo/ws_activities", mode="overwrite")
In the Eventhouse, I added the above table as a shortcut and queried to aggregate by location using geo_info_from_ip_address:
//table name "ws_activities"
external_table("ws_activities")
| extend Location = geo_info_from_ip_address(['Client IP'])
| extend LocationString = strcat_delim(", ",
tostring(Location.city),
tostring(Location.state),
tostring(Location.country))
| extend Latitude = todouble(Location.latitude)
| extend Longitude = todouble(Location.longitude)
| where isnotempty(LocationString) and LocationString != ", , "
| where isnotempty(Latitude) and isnotempty(Longitude)
| summarize EventCount = count() by LocationString, Latitude, Longitude
I live in Portland, OR and recently travelled to Seattle, WA so this checks out:
Rendered the result as a map which shows Portland and Seattle.
The same map is available in Real Time Dashboard so you can track the location of activities in real time and generate alerts if needed!