# Glide AI capabilities
**URL:** https://community.glideapps.com/t/glide-ai-capabilities/79847
**Category:** Ask for Help
**Tags:** ai
**Created:** [February 6, 2025, 6:19pm UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847 "2025-02-06T18:19:54Z")
**Posts on this page:** 11
**Page:** 1
### Author:  [@MattLB](https://community.glideapps.com/u/MattLB)
#### Post date: [February 6, 2025, 6:19pm UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/1 "2025-02-06T18:19:54Z")
Can Glide AI use a table as its input to a prompt? Is this Advanced Reasoning and requires the Enterprise Plan?
I swear that I saw where Glide could ‘reason’ against its own data (rows in a table) but now that I am at my desktop I can NOT find any info - maybe I was dreaming this?
My use case: I have a table which is a feed of activity with users, text column describing the activity, dates, etc and I want to point Glide AI at it. Then use the prompt to find answers or correlations from the description column and the meta data.
I watched/read a lot of videos/articles recently and may have mistaken this capability or maybe, just maybe - this is covered under Advanced Reasoning and is an Enterprise feature.
TIA for what may be a wild goose chase!
---
### Author:  [@MaximeBaker](https://community.glideapps.com/u/MaximeBaker)
#### Post date: [February 6, 2025, 6:34pm UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/2 "2025-02-06T18:34:28Z")
> [@MattLB](#):
>
> Is this Advanced Reasoning and requires the Enterprise Plan?
Yes.
> [@MattLB](#):
>
> I swear that I saw where Glide could ‘reason’ against its own data (rows in a table) but now that I am at my desktop I can NOT find any info - maybe I was dreaming this?
I also think I saw this too! But not able to find it.
---
### Author:  [@MattLB](https://community.glideapps.com/u/MattLB)
#### Post date: [February 6, 2025, 6:37pm UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/3 "2025-02-06T18:37:26Z")
I read this from the docs [https://www.glideapps.com/docs/automation/ai:](https://www.glideapps.com/docs/automation/ai:)
You can start from scratch with a new data table, or use your existing data in input fields to combine _your dataset_ with the power of Glide AI.
So I ‘assumed’ existing data included queries - aka ‘your dataset’.
I am now trying to find more information on Advanced Reasoning with little luck.
Thanks for looking into this.
---
### Author:  [@MattLB](https://community.glideapps.com/u/MattLB)
#### Post date: [February 6, 2025, 6:41pm UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/4 "2025-02-06T18:41:35Z")
But what is funny is this response from my search for more information on Glide AI: “AI results unavailable” 🤣
---
### Author:  [@MaximeBaker](https://community.glideapps.com/u/MaximeBaker)
#### Post date: [February 6, 2025, 6:41pm UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/5 "2025-02-06T18:41:56Z")
Based on this:
> [@Glide AI and sources](https://community.glideapps.com/t/glide-ai-and-sources/79843/5):
>
> Why not use a joined list to retrieve all the JSON data from the query, assuming you have already created a JSON column in the target table? You could then use a template column to concatenate “[” and “]” around the joined list. This final column can be utilized in a Glide AI column to execute the instructions you wanted on this data.
You could also include that in the instruction of the advanced reasoning.
---
### Author:  [@ThinhDinh](https://community.glideapps.com/u/ThinhDinh)
#### Post date: [February 7, 2025, 1:23am UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/6 "2025-02-07T01:23:33Z")
Your method should work, I use it when I don’t have a lot of data.
However, if the context is too big, I would love something easier to work with. I will try Pinecone, but many of my use cases would involve aggregating data (e.g: What is the sum of sales last month), and I’m not sure Pinecone is good for that.
---
### Author:  [@MaximeBaker](https://community.glideapps.com/u/MaximeBaker)
#### Post date: [March 8, 2025, 1:41am UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/7 "2025-03-08T01:41:48Z")
@MattLB I found it !
---
### Author:  [@ThinhDinh](https://community.glideapps.com/u/ThinhDinh)
#### Post date: [March 8, 2025, 3:44am UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/8 "2025-03-08T03:44:04Z")
This is only semi-useful if you want to extract info from X specific rows (I believe X is 1). Most of the time when people complain about this, it’s when they want aggregating data from multiple rows.
---
### Author:  [@MaximeBaker](https://community.glideapps.com/u/MaximeBaker)
#### Post date: [March 8, 2025, 3:46am UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/9 "2025-03-08T03:46:22Z")
I am not sure to follow you…
---
### Author:  [@ThinhDinh](https://community.glideapps.com/u/ThinhDinh)
#### Post date: [March 8, 2025, 3:51am UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/10 "2025-03-08T03:51:20Z")
Say if you have a database of 1000 records about fruit sales, if you ask how many apples you sell at your California store, on March 7th 2025, it might give you the right answer.
If you ask how many apple sales you have in January 2025, it would hallucinate 75% of the time.
It’s not good for aggregating data.
If it’s for asking questions about internal SOP, then yeah, it might be good.
---
### Author:  [@ThinhDinh](https://community.glideapps.com/u/ThinhDinh)
#### Post date: [March 8, 2025, 3:52am UTC](https://community.glideapps.com/t/glide-ai-capabilities/79847/11 "2025-03-08T03:52:23Z")
> [@robot OpenAI Integration — Full Build Using 10+ OpenAI Actions/Columns](https://community.glideapps.com/t/openai-integration-full-build-using-10-openai-actions-columns/62386/22):
>
> Regarding the “Answer question about the table”: It might help to understand the mechanism - the action pulls in one row of data most relevant to the question given and use that as context to answer the question. For example, you have a list of FAQ with the column question and answer - and present it as a front desk helper. Currently not set up to do rollup/aggregation over many rows like Roberts use case. It tends to do well when the columns contain more text like descriptions than categoric…