Copilot ValueError: Schema healthcare1_msft_bronze.substancedefinition is too long to be added in NB

[sharp1FL]

I am trying to use Copilot in a notebook in my HLS public preview workspace.  I get the following error when I execute the Copilot package installation cell.  Any advice on how I can address this?

Thanks.

ValueError: Schema healthcare1_msft_bronze.substancedefinition is too long to be added to the vector store.

[Nabha-Ahmed]

Hi @sharp1FL 

This error usually happens when the schema or table name is too long for the Copilot vector store.
You can try one of the following approaches to fix it:

1. Use a shorter alias or view name

CREATE VIEW bronze_substance AS
SELECT * FROM healthcare1_msft_bronze.substancedefinition

Then reference it in your notebook as:

df = spark.table("bronze_substance")

2. Select only the needed columns

df = spark.table("healthcare1_msft_bronze.substancedefinition").select("column1", "column2")

3. Create a temporary view with a shorter name

spark.table("healthcare1_msft_bronze.substancedefinition").createOrReplaceTempView("substancedef_short")

Then use the short name (substancedef_short) when calling Copilot.
If the issue persists, make sure you are on the latest Fabric Copilot public preview build — some early preview versions have naming length limits.

Hope this helps!

[Akash_Varuna]

Hi @sharp1FL  , Could you please try these steps

• Shorten the schema name to reduce its length.
• Use aliases or map verbose names to shorter ones.
• Preprocess the schema name to truncate it before adding it to the vector store.
• Check if the vector store allows increasing the maximum length limit.
• Split the schema into smaller components (e.g., database, schema, table).
  if this post helped please do give a kudos and accept this as a solution 
  Thanks in Advance