Scaling AI is about more than just better models—it's about better architecture. 🛠️
Salesforce Engineering just released a great breakdown on scaling their AI Engagement Agents to handle 1M+ actions monthly through intelligent dispatching and parallel execution. It’s a masterclass in building for reliability and burst-tolerance in the enterprise era.
Full story: Scaling Sales Agents: Engineering Next-Gen AI:
https://lnkd.in/gNbrkH_8
Scaling Agentforce Apps: Agentforce Ask Me Anything May 2025
Join the next AMA to discover how developers can build scalable Agentforce solutions by testing their implementations for scale, monitoring key metrics, detecting anti-patterns and optimizing workflows.
Date: Wednesday May 28, 2025
Time: 9am PT
Register here: https://www.linkedin.com/events/7320853906807209984/comments/
***Change the Scale of Y-Axis in a Bar Chart***
We are using a Bar Chart to show Investments in each year (refer attachment). We would like the Y-Axis scale to show Amount in Millions only. We do not want to see 500K or 300K in the Y-Axis.
Can we change the starting value and increments to be used in the Scale? Is there an option to change the scale of the Axes?
Example: Based on the data being shown, a scale starting from 1Mil and increment by 1Mil or A scale starting 10Mil and increment by 5Mil
To change the axis starting/ending values: Formatting > Y-Axis (Left) > Add Custom Domain.
I believe the intervals are set dynamically based on the domain/scale; not sure if you can influence this.
Hi all,
I have noticed the revenue figures do not seem to match our external data once i have loaded them into Einstein Analytics via the External Data API, and i think it could be something to do with the Scale and Precision.
The part i' unsure about is that the length and amount of decimal places are not the same on every row of our data, so would i just use the Scale and Precision for the largest ?
If you see a different result for grouped/ungrouped, this usually means nulls in the grouping. Raw dataset is to make sure a dataflow didn't need to run.
Have you compared the row count of the source data and what's in EA? If EA's count of rows is lower, something may have gone wrong with the dataload, in which case I'd suggest trying again.