---
id: video/fgWFOMPJd6U
type: video
title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
summary: How Summarize with Copilot in Business Central works. A multi-prompt pipeline finds, scores and summarizes insights from page fields, related statistics pages and fact boxes, using the user's role as context. The video also covers how the team tests AI features (harm, grounding, functional checks), the migration of the summarization prompts to GPT-4.1 mini, and how AL developers can improve summaries with metadata, fact boxes and tooltips.
tier: official
language: en
tags:
  - summarization with copilot
  - data-driven insights
  - multi-prompt pipeline
  - user context
  - related entities
  - fact boxes
  - statistics pages
  - user intent
  - telemetry
  - prompt engineering
  - ai testing
  - grounding
system: copilot
review:
  state: reviewed
  by: opus
  at: "2026-10-07T22:52:16.435Z"
  flags: []
generated:
  at: "2026-10-07T22:52:16.488Z"
  pipeline: 0.2.0
  prompts:
    extract-video: 1
    summarize-video: 2
  input_hash: cfaa1e67330f826474a92e78cc3014deacc03295c354817497b808e8c3c411ab
evidence:
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=814s
    title: "External Data Integration for Summarization: announced"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 814
    quote: in future we do have plans of increasing both the tracks the user context track but also the data context
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=103s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 103
    quote: So summarizing with copilot highlights what's most important and urgent for the user on almost every business central page.
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=152s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 152
    quote: So summarization with copilot is available on any business central uh single entity pages like card documents or list plus.
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=232s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 232
    quote: The way we build summarization, it's pretty much a three-step process.
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=476s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 476
    quote: we filter out only the pages that are named statistics because that's where the aggregated data exist and that's very good for summarization
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=531s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 531
    quote: So we started with sending pretty much uh all the ledger entries and all and we soon realized uh it's not very performant cuz
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=814s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 814
    quote: in future we do have plans of increasing both the tracks the user context track but also the data context.
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=857s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 857
    quote: there's definitely very good early adoption. We are seeing that. Um uh there's a little bit of lesser percentage of people clicking show more.
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=974s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 974
    quote: what I learned more is that tests are super super important especially because you change one line of prompt and the response could be
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1129s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1129
    quote: Uh definitely one of the new things about uh testing with AI features is that you use AI to test AI features.
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1287s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1287
    quote: So we came up with some tremendous large test data set containing multiple roles and multiple scenarios with different entities. uh some for example
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1519s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1519
    quote: And now uh we have migrated all of them to 41 mini.
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1556s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1556
    quote: Uh it depends on the need of the feature, right? Definitely. uh for for summarization specifically um we believe that 41 mini is enough
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1590s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1590
    quote: Some of the prompts, for example, the scoring part could be probably better done with uh other models than 41 mini uh because they
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1627s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1627
    quote: do try it with other models do and that's why have a good set of test so you could easily switch between models
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1647s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1647
    quote: Another advantage of small language or small language mod is that they're also cheaper, right? Definitely. Yes. But you need to maintain them yourself.
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1684s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1684
    quote: so the as i said summarization works ex out of the box with custom entities
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1697s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1697
    quote: Right now it is not directly extensible from AL in the sense that uh they they cannot provide for example more data sources or
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1765s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1765
    quote: as a minimum put good tool tips on the on the pages then you'll get probably a better summary because that
  - kind: video
    url: https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1781s
    title: "Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot"
    date: "2025-10-16T14:01:04.000Z"
    commit: null
    t: 1781
    quote: think of AI as your co-pilot. There's a reason we call it co-pilot because it is uh one of them, right?
links:
  learn: []
  objects: []
  features: []
  topics: []
  localizations: []
  videos: []
  posts: []
  guidelines: []
video_id: fgWFOMPJd6U
channel: yt-microsoft
source_name: Microsoft Dynamics 365 Business Central (YouTube)
url: https://www.youtube.com/watch?v=fgWFOMPJd6U
published_at: "2025-10-16T14:01:04.000Z"
duration_s: 1832
captions: full
audience:
  - developer
  - partner
  - functional consultant
chapters:
  - t: 0
    title: Introduction to AI Experiences and Summarize with Copilot
  - t: 103
    title: "Feature Overview: What Summarize with Copilot Does"
  - t: 185
    title: "Under the Hood: Data Collection and Insights Generation"
  - t: 276
    title: "Multi-Prompt Pipeline: Find Insights and Summarize"
  - t: 372
    title: Data Source Selection and Related Entities
  - t: 508
    title: "Evolution of Summarization: From Simple to Data-Driven"
  - t: 683
    title: "Future Roadmap: User Intent and Enhanced Context"
  - t: 814
    title: Telemetry and Show More Feature
  - t: 895
    title: Show more feature and chat experience
  - t: 944
    title: Prompt engineering and architecture for AI
  - t: 1034
    title: Testing strategy for AI features
  - t: 1150
    title: Grounding and accuracy testing methods
  - t: 1239
    title: Functional testing and quality metrics
  - t: 1374
    title: Customer feedback and experimentation
  - t: 1519
    title: Model selection and migration strategy
  - t: 1647
    title: Small language models and extensibility
  - t: 1780
    title: Guidance for Partners, Developers and Customers
features:
  - name: Summarize with Copilot
    status: unclear
    t: 103
    verified: false
    status_source: video
  - name: Multi-Prompt Pipeline for Summarization
    status: unclear
    t: 232
    verified: false
    status_source: video
  - name: Statistics Page Integration for Summarization
    status: unclear
    t: 451
    verified: false
    status_source: video
  - name: User Role-Based Filtering for Insights
    status: unclear
    t: 288
    verified: false
    status_source: video
  - name: User Intent Detection for Future Summarization
    status: unclear
    t: 723
    verified: false
    status_source: video
  - name: External Data Integration for Summarization
    status: announced
    t: 814
    verified: true
    status_source: video
  - name: Harm testing for AI responses
    status: unclear
    t: 1054
    verified: false
    status_source: video
  - name: Accuracy and grounding testing
    status: unclear
    t: 1139
    verified: false
    status_source: video
  - name: Functional testing for summarization
    status: unclear
    t: 1239
    verified: false
    status_source: video
  - name: Summary quality metrics
    status: unclear
    t: 1324
    verified: false
    status_source: video
  - name: Experimentation framework for AI features
    status: unclear
    t: 1396
    verified: false
    status_source: video
  - name: Small language model advantages and trade-offs
    status: unclear
    t: 1647
    verified: false
    status_source: video
  - name: Summarization with multiple model support
    status: unclear
    t: 1576
    verified: false
    status_source: video
  - name: Summarization with custom entities
    status: unclear
    t: 1684
    verified: false
    status_source: video
  - name: Summary metadata and AI context improvements
    status: unclear
    t: 1712
    verified: false
    status_source: video
objects_mentioned: []
quotes:
  - t: 103
    text: So summarizing with copilot highlights what's most important and urgent for the user on almost every business central page.
    check: exact
  - t: 152
    text: So summarization with copilot is available on any business central uh single entity pages like card documents or list plus.
    check: exact
  - t: 232
    text: The way we build summarization, it's pretty much a three-step process.
    check: exact
  - t: 476
    text: we filter out only the pages that are named statistics because that's where the aggregated data exist and that's very good for summarization
    check: fuzzy
  - t: 531
    text: So we started with sending pretty much uh all the ledger entries and all and we soon realized uh it's not very performant cuz
    check: exact
  - t: 814
    text: in future we do have plans of increasing both the tracks the user context track but also the data context.
    check: exact
  - t: 857
    text: there's definitely very good early adoption. We are seeing that. Um uh there's a little bit of lesser percentage of people clicking show more.
    check: exact
  - t: 974
    text: what I learned more is that tests are super super important especially because you change one line of prompt and the response could be
    check: exact
  - t: 1129
    text: Uh definitely one of the new things about uh testing with AI features is that you use AI to test AI features.
    check: exact
  - t: 1287
    text: So we came up with some tremendous large test data set containing multiple roles and multiple scenarios with different entities. uh some for example
    check: exact
  - t: 1519
    text: And now uh we have migrated all of them to 41 mini.
    check: exact
  - t: 1556
    text: Uh it depends on the need of the feature, right? Definitely. uh for for summarization specifically um we believe that 41 mini is enough
    check: exact
  - t: 1590
    text: Some of the prompts, for example, the scoring part could be probably better done with uh other models than 41 mini uh because they
    check: exact
  - t: 1627
    text: do try it with other models do and that's why have a good set of test so you could easily switch between models
    check: exact
  - t: 1647
    text: Another advantage of small language or small language mod is that they're also cheaper, right? Definitely. Yes. But you need to maintain them yourself.
    check: exact
  - t: 1684
    text: so the as i said summarization works ex out of the box with custom entities
    check: fuzzy
  - t: 1697
    text: Right now it is not directly extensible from AL in the sense that uh they they cannot provide for example more data sources or
    check: exact
  - t: 1765
    text: as a minimum put good tool tips on the on the pages then you'll get probably a better summary because that
    check: fuzzy
  - t: 1781
    text: think of AI as your co-pilot. There's a reason we call it co-pilot because it is uh one of them, right?
    check: exact
---

# Business Central Under the Hood 11: New Data-Driven Copilot Experiences – Summarize with Copilot

> How Summarize with Copilot in Business Central works. A multi-prompt pipeline finds, scores and summarizes insights from page fields, related statistics pages and fact boxes, using the user's role as context. The video also covers how the team tests AI features (harm, grounding, functional checks), the migration of the summarization prompts to GPT-4.1 mini, and how AL developers can improve summaries with metadata, fact boxes and tooltips.

[Watch on YouTube](https://www.youtube.com/watch?v=fgWFOMPJd6U) · Microsoft Dynamics 365 Business Central (YouTube) · 2025-10-16 · 30:32 · tier official · reviewed (checked by Opus)

## Overview

Episode 11 of Business Central Under the Hood explains how Summarize with Copilot works. It runs on single entity pages such as card, document and list pages, and shows the top 2-3 insights in a summary fact box, with a Show more option that opens a chat experience. Behind it is a pipeline: a find insights prompt reads data from the current page, related pages and fact boxes in parallel, then a summarize prompt picks the top insights and applies formatting.

The second half covers engineering practice: testing AI with AI, grounding and harm checks, quality metrics, experimentation, model selection and migration to smaller models. It closes with guidance for partners and developers on custom entities, where summarization works out of the box but cannot yet be extended directly from AL.

## Key points

- Summarize with Copilot is available on any single entity page (card, document, ListPlus), including custom partner pages, and appears in a summary fact box at the top of the fact box pane.
- The pipeline has a find insights prompt (current page fields, related pages and fact boxes, run in parallel) followed by a summarize prompt that selects the top-scored insights and formats them in markdown.
- Related pages are actions on the current page that lead to pages named statistics. Aggregated data performed better than sending raw ledger entries.
- Role is the current user context signal used to prioritize insights. Inferring user intent from navigation patterns and frequently used pages is planned, as are more data sources.
- Insights must reference source data IDs via function calling, which supports grounding. Testing also covers harm checks and functional checks with large role and scenario datasets, plus metrics like brevity and precision.
- Summarization prompts moved from 4o and 4o mini to 4.1 mini for speed. The scoring prompt might benefit from reasoning models such as o1, and the team is still experimenting.
- Developers cannot yet add data sources or override summary text from AL (work in progress). They can improve results with metadata, extra fact boxes, actions to statistics pages, and clear captions and tooltips.

## Chapters

- [0:00](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=0s) Introduction to AI Experiences and Summarize with Copilot
- [1:43](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=103s) Feature Overview: What Summarize with Copilot Does
- [3:05](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=185s) Under the Hood: Data Collection and Insights Generation
- [4:36](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=276s) Multi-Prompt Pipeline: Find Insights and Summarize
- [6:12](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=372s) Data Source Selection and Related Entities
- [8:28](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=508s) Evolution of Summarization: From Simple to Data-Driven
- [11:23](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=683s) Future Roadmap: User Intent and Enhanced Context
- [13:34](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=814s) Telemetry and Show More Feature
- [14:55](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=895s) Show more feature and chat experience
- [15:44](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=944s) Prompt engineering and architecture for AI
- [17:14](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1034s) Testing strategy for AI features
- [19:10](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1150s) Grounding and accuracy testing methods
- [20:39](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1239s) Functional testing and quality metrics
- [22:54](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1374s) Customer feedback and experimentation
- [25:19](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1519s) Model selection and migration strategy
- [27:27](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1647s) Small language models and extensibility
- [29:40](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1780s) Guidance for Partners, Developers and Customers

## Features

| Feature | Status | At | Evidence |
|---|---|---|---|
| Summarize with Copilot | status not stated, demoed | [1:43](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=103s) |  |
| Multi-Prompt Pipeline for Summarization | status not stated | [3:52](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=232s) |  |
| Statistics Page Integration for Summarization | status not stated | [7:31](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=451s) |  |
| User Role-Based Filtering for Insights | status not stated | [4:48](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=288s) |  |
| User Intent Detection for Future Summarization | status not stated | [12:03](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=723s) |  |
| External Data Integration for Summarization | announced | [13:34](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=814s) | "in future we do have plans of increasing both the tracks the user context track but also the data context" ([13:34](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=814s)) |
| Harm testing for AI responses | status not stated | [17:34](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1054s) |  |
| Accuracy and grounding testing | status not stated | [18:59](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1139s) |  |
| Functional testing for summarization | status not stated | [20:39](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1239s) |  |
| Summary quality metrics | status not stated | [22:04](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1324s) |  |
| Experimentation framework for AI features | status not stated | [23:16](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1396s) |  |
| Small language model advantages and trade-offs | status not stated | [27:27](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1647s) |  |
| Summarization with multiple model support | status not stated | [26:16](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1576s) |  |
| Summarization with custom entities | status not stated | [28:04](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1684s) |  |
| Summary metadata and AI context improvements | status not stated | [28:32](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1712s) |  |

## Quotes

- [1:43](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=103s) "So summarizing with copilot highlights what's most important and urgent for the user on almost every business central page."
- [2:32](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=152s) "So summarization with copilot is available on any business central uh single entity pages like card documents or list plus."
- [3:52](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=232s) "The way we build summarization, it's pretty much a three-step process."
- [7:56](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=476s) "we filter out only the pages that are named statistics because that's where the aggregated data exist and that's very good for summarization"
- [8:51](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=531s) "So we started with sending pretty much uh all the ledger entries and all and we soon realized uh it's not very performant cuz"
- [13:34](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=814s) "in future we do have plans of increasing both the tracks the user context track but also the data context."
- [14:17](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=857s) "there's definitely very good early adoption. We are seeing that. Um uh there's a little bit of lesser percentage of people clicking show more."
- [16:14](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=974s) "what I learned more is that tests are super super important especially because you change one line of prompt and the response could be"
- [18:49](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1129s) "Uh definitely one of the new things about uh testing with AI features is that you use AI to test AI features."
- [21:27](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1287s) "So we came up with some tremendous large test data set containing multiple roles and multiple scenarios with different entities. uh some for example"
- [25:19](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1519s) "And now uh we have migrated all of them to 41 mini."
- [25:56](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1556s) "Uh it depends on the need of the feature, right? Definitely. uh for for summarization specifically um we believe that 41 mini is enough"
- [26:30](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1590s) "Some of the prompts, for example, the scoring part could be probably better done with uh other models than 41 mini uh because they"
- [27:07](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1627s) "do try it with other models do and that's why have a good set of test so you could easily switch between models"
- [27:27](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1647s) "Another advantage of small language or small language mod is that they're also cheaper, right? Definitely. Yes. But you need to maintain them yourself."
- [28:04](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1684s) "so the as i said summarization works ex out of the box with custom entities"
- [28:17](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1697s) "Right now it is not directly extensible from AL in the sense that uh they they cannot provide for example more data sources or"
- [29:25](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1765s) "as a minimum put good tool tips on the on the pages then you'll get probably a better summary because that"
- [29:41](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1781s) "think of AI as your co-pilot. There's a reason we call it co-pilot because it is uh one of them, right?"

## Disclaimers in the video

- [12:03](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=723s) other: In future we do have plans of increasing both the tracks the user context track but also the data context.
- [13:34](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=814s) preview: we also want to stabilize uh more and more with the current version and and then in future we do have plans of increasing both the tracks the user context track but also the data context
- [13:47](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=827s) coming-later: We could easily think of uh you know integrating with external data be it um data from LinkedIn or any other external sources
- [23:37](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1417s) coming-later: we're working um uh on doing what we call experimentation. I don't want to tell much about it because that's that could be a subject an entire an entire episode
- [26:30](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1590s) preview: Some of the prompts, for example, the scoring part could be probably better done with uh other models than 41 mini uh because they might have better reasoning capabilities and we are still figuring out
- [28:17](https://www.youtube.com/watch?v=fgWFOMPJd6U&t=1697s) coming-later: Right now it is not directly extensible from AL in the sense that uh they they cannot provide for example more data sources or they cannot override the text that they see in the summary in one way or another yet but we are working on it.

Presenters (as heard): Vincent, Monica.