---
id: video/ytYKvn2MiGs
type: video
title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
summary: "Extending Copilot in Business Central with AL code, 2024 release wave 1: prompt dialog instructional text and prompt guides, Copilot actions on lists, token counting in the AOI Token codeunit, and function calling with tool choice. Also covers how Microsoft keeps models current with latest/preview versions and the announced managed, shared AI resources with AI credit governance (private preview in 24.x, public preview in 25.x, GA TBD). Ends with partner Q&A on data privacy, using your own data via prompts, and choosing GPT-3.5 Turbo where it fits."
tier: official
language: en
tags:
  - prompt dialog
  - token counting
  - function calling
  - copilot actions
  - ai module
  - azure openai
  - llm
  - al development
  - copilot capabilities
  - prompt guides
  - model versioning
  - tool choice
system: copilot
review:
  state: reviewed
  by: opus
  at: "2026-10-07T23:09:05.749Z"
  flags: []
generated:
  at: "2026-10-07T23:09:05.803Z"
  pipeline: 0.2.0
  prompts:
    extract-video: 1
    summarize-video: 2
  input_hash: 3dcb6c70379752c2c7562a60dcc2f2fd9184d22b320dc5ce5f17daa9e4b0967a
evidence:
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1573s
    title: "Managed and Shared AI Resources: preview"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1573
    quote: private preview will happen pretty soon after the business centrer launch event
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1573s
    title: "Capacity-based resource governance and pricing: announced"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1573
    quote: capacity based resource governance and pricing will be modeled and finalized through simulated limit and quota enforcement in private or public preview
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1634s
    title: "Limit and quota publishing and enforcement: announced"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1634
    quote: in public preview we will also make sure that the proper limit will be published and enforced while the proper quota might also be
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=171s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 171
    quote: now we can provide concise instructions to help users get started they will be displayed
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=341s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 341
    quote: token accounting allows you to count the number of tokens that will be used by your input before sending a request to Azure open
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=416s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 416
    quote: you would like to evaluate how many tokens are going to be used because then you can estimate what your potential cost is um
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=436s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 436
    quote: a model has a limit to how many tokens there are uh that it can actually take in as input and now if you
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=476s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 476
    quote: agnostic so whether you're using um gpt 4 uh january version or gpt 4 june version it does not matter
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=517s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 517
    quote: the input is of type secret text and why is that uh and that the reason for that is because your input we consider
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=574s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 574
    quote: it allows you to describe functions to the assistant model uh and have it return uh the functions that should be called along with
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=833s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 833
    quote: you should always retrieve that from isolated storage or Azure key volt for example uh because your description as I had mentioned these we
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=995s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 995
    quote: if you use it with functions the response can either be a function or it can be a regular chat response so you do
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1015s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1015
    quote: if you set a specific Tool uh specific function it will guarantee that response will always call that function however if your user inputs
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1076s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1076
    quote: we keep two versions of each model available latest and preview not that we don't really specify if the latest is the June model
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1289s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1289
    quote: when it's time to bring this AI infused solution into production at scale especially when they are combining multiple copilot features built by different
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1370s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1370
    quote: partners and isv can simply focus on satisfying responsible AI requirement to access our shared resources building their calet features based on their intellectual
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1573s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1573
    quote: capacity based resource governance and pricing will be modeled and finalized through simulated limit and quota enforcement in private or public preview
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1613s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1613
    quote: this 24x release so either 24.0 in April or 24.1 in May right next public preview we will introduce a new process to satisfy
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1613s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1613
    quote: the ETA for priate review is pretty soon so this 24x release so either 24.0 in April or 24.1 in May
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1655s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1655
    quote: General availability is when the proper limits and qu will be both published and enforced the ETA still to be determined or TBD depending
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1655s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1655
    quote: so the ETF for public preview will be the next release around 25th uh 25x version
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1797s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1797
    quote: which business Central data does Asia open AI have access to the answer is none our commitment is that your data is your data
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1829s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1829
    quote: what if I need to give my data to the model how can I use my data with aure open Ai and the answer
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1890s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 1890
    quote: our recommendation is to use GPT 3.5 turbo whenever it fits your needs because it's faster first of all it's also cheaper for you
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=2027s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 2027
    quote: the AL part of our toolkit is actually open source is part of our aka.ms /bc apps Repro in GitHub and you can contribute
  - kind: video
    url: https://www.youtube.com/watch?v=ytYKvn2MiGs&t=2047s
    title: "What's New: Extending Copilot Using AL Code (2024 Release Wave 1)"
    date: "2024-04-04T13:30:33.000Z"
    commit: null
    t: 2047
    quote: Microsoft cloud runs on trust AI is no exception your data is only yours it's your data the AI models are not trained on
links:
  learn: []
  objects: []
  features: []
  topics: []
  localizations: []
  videos: []
  posts: []
  guidelines: []
video_id: ytYKvn2MiGs
channel: yt-microsoft
source_name: Microsoft Dynamics 365 Business Central (YouTube)
url: https://www.youtube.com/watch?v=ytYKvn2MiGs
published_at: "2024-04-04T13:30:33.000Z"
duration_s: 2169
captions: full
audience:
  - developer
  - partner
  - functional consultant
  - decision maker
chapters:
  - t: 0
    title: Developer Tools Recap and AI Module Overview
  - t: 151
    title: Prompt Dialog Enhancements
  - t: 281
    title: Copilot Actions and Implementation
  - t: 341
    title: Token Counting Fundamentals
  - t: 456
    title: Token Counting Methods and Implementation
  - t: 553
    title: Function Calling Concepts and Scenarios
  - t: 664
    title: Function Calling Technical Details
  - t: 793
    title: Function Calling Implementation in AL
  - t: 974
    title: Tool Choice Modes and Validation
  - t: 1035
    title: Keeping Current with Model Versions
  - t: 1154
    title: Model Version Rollout Process and Managed AI Resources
  - t: 1370
    title: AI Credit System and Resource Governance
  - t: 1573
    title: Roadmap and Timeline for AI Resources
  - t: 1675
    title: Developer Toolkit, Data Privacy and Resources
features:
  - name: Instructional Text in Prompt Dialog
    status: unclear
    t: 171
    verified: false
    status_source: video
  - name: Prompt Guides
    status: unclear
    t: 212
    verified: false
    status_source: video
  - name: Copilot Actions on Lists and Dialogs
    status: unclear
    t: 281
    verified: false
    status_source: video
  - name: Token Counting
    status: unclear
    t: 341
    verified: false
    status_source: video
  - name: Token Counting Methods in AOI Token Code Unit
    status: unclear
    t: 456
    verified: false
    status_source: video
  - name: Function Calling
    status: unclear
    t: 553
    verified: false
    status_source: video
  - name: Function Calling in Copilot
    status: unclear
    t: 813
    verified: false
    status_source: video
  - name: Function Calling Tool Choice
    status: unclear
    t: 853
    verified: false
    status_source: video
  - name: Tool Choice Auto Mode
    status: unclear
    t: 974
    verified: false
    status_source: video
  - name: Tool Choice Specific Function Mode
    status: unclear
    t: 1015
    verified: false
    status_source: video
  - name: Model Version Management
    status: unclear
    t: 1096
    verified: false
    status_source: video
  - name: Version Agnostic Model Strings
    status: unclear
    t: 1116
    verified: false
    status_source: video
  - name: Managed and Shared AI Resources
    status: preview
    t: 1254
    verified: true
    status_source: video
  - name: AI Credit Resource Governance
    status: unclear
    t: 1451
    verified: false
    status_source: video
  - name: Operational Limits for Fair Use
    status: unclear
    t: 1491
    verified: false
    status_source: video
  - name: AI Quota Entitlements
    status: unclear
    t: 1512
    verified: false
    status_source: video
  - name: Capacity-based resource governance and pricing
    status: announced
    t: 1573
    verified: true
    status_source: video
  - name: Responsible AI process for non-OpenAI accounts
    status: unclear
    t: 1634
    verified: false
    status_source: video
  - name: Limit and quota publishing and enforcement
    status: announced
    t: 1634
    verified: true
    status_source: video
  - name: Developer toolkit for Copilot in Business Central
    status: unclear
    t: 1696
    verified: false
    status_source: video
  - name: Azure OpenAI data access control
    status: unclear
    t: 1797
    verified: false
    status_source: video
  - name: Data source capability for Azure OpenAI
    status: unclear
    t: 1870
    verified: false
    status_source: video
  - name: GPT-3.5 Turbo model recommendation
    status: unclear
    t: 1890
    verified: false
    status_source: video
  - name: AL toolkit open source on GitHub
    status: unclear
    t: 2027
    verified: false
    status_source: video
  - name: Data privacy and AI model training
    status: unclear
    t: 2047
    verified: false
    status_source: video
objects_mentioned:
  - codeunit AOI Token
  - codeunit get weather co-pilot function
  - other get weather co-pilot function
  - other get current weather
  - other Azure OpenAI data sources capability
  - other GPT 3.5 turbo model
  - other GPT 4 model
  - other GPT 4 Turbo model
quotes:
  - t: 171
    text: now we can provide concise instructions to help users get started they will be displayed
    check: fuzzy
  - t: 341
    text: token accounting allows you to count the number of tokens that will be used by your input before sending a request to Azure open
    check: exact
  - t: 416
    text: you would like to evaluate how many tokens are going to be used because then you can estimate what your potential cost is um
    check: exact
  - t: 436
    text: a model has a limit to how many tokens there are uh that it can actually take in as input and now if you
    check: snapped
  - t: 476
    text: agnostic so whether you're using um gpt 4 uh january version or gpt 4 june version it does not matter
    check: fuzzy
  - t: 517
    text: the input is of type secret text and why is that uh and that the reason for that is because your input we consider
    check: exact
  - t: 574
    text: it allows you to describe functions to the assistant model uh and have it return uh the functions that should be called along with
    check: exact
  - t: 833
    text: you should always retrieve that from isolated storage or Azure key volt for example uh because your description as I had mentioned these we
    check: exact
  - t: 995
    text: if you use it with functions the response can either be a function or it can be a regular chat response so you do
    check: exact
  - t: 1015
    text: if you set a specific Tool uh specific function it will guarantee that response will always call that function however if your user inputs
    check: exact
  - t: 1076
    text: we keep two versions of each model available latest and preview not that we don't really specify if the latest is the June model
    check: snapped
  - t: 1289
    text: when it's time to bring this AI infused solution into production at scale especially when they are combining multiple copilot features built by different
    check: exact
  - t: 1370
    text: partners and isv can simply focus on satisfying responsible AI requirement to access our shared resources building their calet features based on their intellectual
    check: exact
  - t: 1573
    text: capacity based resource governance and pricing will be modeled and finalized through simulated limit and quota enforcement in private or public preview
    check: exact
  - t: 1613
    text: this 24x release so either 24.0 in April or 24.1 in May right next public preview we will introduce a new process to satisfy
    check: snapped
  - t: 1613
    text: the ETA for priate review is pretty soon so this 24x release so either 24.0 in April or 24.1 in May
    check: exact
  - t: 1655
    text: General availability is when the proper limits and qu will be both published and enforced the ETA still to be determined or TBD depending
    check: exact
  - t: 1655
    text: so the ETF for public preview will be the next release around 25th uh 25x version
    check: exact
  - t: 1797
    text: which business Central data does Asia open AI have access to the answer is none our commitment is that your data is your data
    check: exact
  - t: 1829
    text: what if I need to give my data to the model how can I use my data with aure open Ai and the answer
    check: exact
  - t: 1890
    text: our recommendation is to use GPT 3.5 turbo whenever it fits your needs because it's faster first of all it's also cheaper for you
    check: exact
  - t: 2027
    text: the AL part of our toolkit is actually open source is part of our aka.ms /bc apps Repro in GitHub and you can contribute
    check: exact
  - t: 2047
    text: Microsoft cloud runs on trust AI is no exception your data is only yours it's your data the AI models are not trained on
    check: snapped
---

# What's New: Extending Copilot Using AL Code (2024 Release Wave 1)

> Extending Copilot in Business Central with AL code, 2024 release wave 1: prompt dialog instructional text and prompt guides, Copilot actions on lists, token counting in the AOI Token codeunit, and function calling with tool choice. Also covers how Microsoft keeps models current with latest/preview versions and the announced managed, shared AI resources with AI credit governance (private preview in 24.x, public preview in 25.x, GA TBD). Ends with partner Q&A on data privacy, using your own data via prompts, and choosing GPT-3.5 Turbo where it fits.

[Watch on YouTube](https://www.youtube.com/watch?v=ytYKvn2MiGs) · Microsoft Dynamics 365 Business Central (YouTube) · 2024-04-04 · 36:09 · tier official · reviewed (checked by Opus)

## Overview

The session covers new AL capabilities for building Copilot features. It starts with prompt dialog changes (instructional text, prompt guides) and copilot actions on lists, list parts, worksheets and dialogs. It then covers counting input tokens before calling Azure OpenAI, and function calling, including the tool choice modes and the validation each one needs.

The second half covers keeping up with model versions using version-agnostic strings and the latest and preview model versions. It also covers managed and shared AI resources, the AI credit, limits and quotas, and the roadmap, which depends on preview feedback. It ends with data privacy answers, a model recommendation and the open source AL toolkit on GitHub. The demo code is not production ready.

## Key points

- Set the instructional text property on prompt area fields to show general instructions in the prompt dialog. It is not meant for example prompts.
- Prompt guides are prompt templates opened from a book icon on the prompt dialog. They are added as a new prompt guide area in actions on a prompt dialog page and can be grouped.
- Copilot actions on lists, list parts, worksheets and standard dialogs go in a new prompting area in actions.
- Token counting covers input tokens only, not output. Methods in the AOI Token codeunit cover GPT 3.5, GPT 4 and other models, take SecretText input, and do not yet include the system meta prompt.
- Function calling: describe functions to the model and it returns which to call with arguments. Do not hardcode function descriptions in text. Keep them in isolated storage or Azure Key Vault because they are IP.
- Tool choice auto can return either a function call or a normal chat reply, so handle the error case. Setting a specific function guarantees it is called, but you must validate the arguments.
- Two versions of each model are kept, latest and preview. Use version-agnostic strings such as turbo-latest and test your feature, since versions may perform differently.

## Chapters

- [0:00](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=0s) Developer Tools Recap and AI Module Overview
- [2:31](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=151s) Prompt Dialog Enhancements
- [4:41](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=281s) Copilot Actions and Implementation
- [5:41](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=341s) Token Counting Fundamentals
- [7:36](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=456s) Token Counting Methods and Implementation
- [9:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=553s) Function Calling Concepts and Scenarios
- [11:04](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=664s) Function Calling Technical Details
- [13:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=793s) Function Calling Implementation in AL
- [16:14](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=974s) Tool Choice Modes and Validation
- [17:15](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1035s) Keeping Current with Model Versions
- [19:14](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1154s) Model Version Rollout Process and Managed AI Resources
- [22:50](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1370s) AI Credit System and Resource Governance
- [26:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1573s) Roadmap and Timeline for AI Resources
- [27:55](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1675s) Developer Toolkit, Data Privacy and Resources

## Features

| Feature | Status | At | Evidence |
|---|---|---|---|
| Instructional Text in Prompt Dialog | status not stated, demoed | [2:51](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=171s) |  |
| Prompt Guides | status not stated, demoed | [3:32](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=212s) |  |
| Copilot Actions on Lists and Dialogs | status not stated, demoed | [4:41](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=281s) |  |
| Token Counting | status not stated, demoed | [5:41](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=341s) |  |
| Token Counting Methods in AOI Token Code Unit | status not stated, demoed | [7:36](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=456s) |  |
| Function Calling | status not stated, demoed | [9:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=553s) |  |
| Function Calling in Copilot | status not stated, demoed | [13:33](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=813s) |  |
| Function Calling Tool Choice | status not stated, demoed | [14:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=853s) |  |
| Tool Choice Auto Mode | status not stated, demoed | [16:14](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=974s) |  |
| Tool Choice Specific Function Mode | status not stated | [16:55](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1015s) |  |
| Model Version Management | status not stated, demoed | [18:16](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1096s) |  |
| Version Agnostic Model Strings | status not stated, demoed | [18:36](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1116s) |  |
| Managed and Shared AI Resources | preview | [20:54](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1254s) | "private preview will happen pretty soon after the business centrer launch event" ([26:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1573s)) |
| AI Credit Resource Governance | status not stated | [24:11](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1451s) |  |
| Operational Limits for Fair Use | status not stated | [24:51](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1491s) |  |
| AI Quota Entitlements | status not stated | [25:12](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1512s) |  |
| Capacity-based resource governance and pricing | announced | [26:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1573s) | "capacity based resource governance and pricing will be modeled and finalized through simulated limit and quota enforcement in private or public preview" ([26:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1573s)) |
| Responsible AI process for non-OpenAI accounts | status not stated | [27:14](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1634s) |  |
| Limit and quota publishing and enforcement | announced | [27:14](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1634s) | "in public preview we will also make sure that the proper limit will be published and enforced while the proper quota might also be" ([27:14](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1634s)) |
| Developer toolkit for Copilot in Business Central | status not stated | [28:16](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1696s) |  |
| Azure OpenAI data access control | status not stated | [29:57](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1797s) |  |
| Data source capability for Azure OpenAI | status not stated | [31:10](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1870s) |  |
| GPT-3.5 Turbo model recommendation | status not stated | [31:30](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1890s) |  |
| AL toolkit open source on GitHub | status not stated | [33:47](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=2027s) |  |
| Data privacy and AI model training | status not stated | [34:07](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=2047s) |  |

## AL objects mentioned

As heard in the captions. A name that matches one object page by exact type and name links to it; the others stay as named.

- codeunit "AOI Token" at [7:56](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=476s)
- codeunit "get weather co-pilot function" at [13:33](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=813s)
- other "get weather co-pilot function" at [13:33](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=813s)
- other "get current weather" at [14:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=853s)
- other "Azure OpenAI data sources capability" at [29:17](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1757s)
- other "GPT 3.5 turbo model" at [31:30](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1890s)
- other "GPT 4 model" at [31:51](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1911s)
- other "GPT 4 Turbo model" at [31:51](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1911s)

Not found in BC28-30: codeunit "AOI Token", codeunit "get weather co-pilot function".

## Quotes

- [2:51](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=171s) "now we can provide concise instructions to help users get started they will be displayed"
- [5:41](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=341s) "token accounting allows you to count the number of tokens that will be used by your input before sending a request to Azure open"
- [6:56](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=416s) "you would like to evaluate how many tokens are going to be used because then you can estimate what your potential cost is um"
- [7:16](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=436s) "a model has a limit to how many tokens there are uh that it can actually take in as input and now if you"
- [7:56](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=476s) "agnostic so whether you're using um gpt 4 uh january version or gpt 4 june version it does not matter"
- [8:37](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=517s) "the input is of type secret text and why is that uh and that the reason for that is because your input we consider"
- [9:34](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=574s) "it allows you to describe functions to the assistant model uh and have it return uh the functions that should be called along with"
- [13:53](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=833s) "you should always retrieve that from isolated storage or Azure key volt for example uh because your description as I had mentioned these we"
- [16:35](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=995s) "if you use it with functions the response can either be a function or it can be a regular chat response so you do"
- [16:55](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1015s) "if you set a specific Tool uh specific function it will guarantee that response will always call that function however if your user inputs"
- [17:56](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1076s) "we keep two versions of each model available latest and preview not that we don't really specify if the latest is the June model"
- [21:29](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1289s) "when it's time to bring this AI infused solution into production at scale especially when they are combining multiple copilot features built by different"
- [22:50](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1370s) "partners and isv can simply focus on satisfying responsible AI requirement to access our shared resources building their calet features based on their intellectual"
- [26:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1573s) "capacity based resource governance and pricing will be modeled and finalized through simulated limit and quota enforcement in private or public preview"
- [26:53](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1613s) "this 24x release so either 24.0 in April or 24.1 in May right next public preview we will introduce a new process to satisfy"
- [26:53](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1613s) "the ETA for priate review is pretty soon so this 24x release so either 24.0 in April or 24.1 in May"
- [27:35](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1655s) "General availability is when the proper limits and qu will be both published and enforced the ETA still to be determined or TBD depending"
- [27:35](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1655s) "so the ETF for public preview will be the next release around 25th uh 25x version"
- [29:57](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1797s) "which business Central data does Asia open AI have access to the answer is none our commitment is that your data is your data"
- [30:29](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1829s) "what if I need to give my data to the model how can I use my data with aure open Ai and the answer"
- [31:30](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1890s) "our recommendation is to use GPT 3.5 turbo whenever it fits your needs because it's faster first of all it's also cheaper for you"
- [33:47](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=2027s) "the AL part of our toolkit is actually open source is part of our aka.ms /bc apps Repro in GitHub and you can contribute"
- [34:07](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=2047s) "Microsoft cloud runs on trust AI is no exception your data is only yours it's your data the AI models are not trained on"

## Disclaimers in the video

- [8:37](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=517s) coming-later: these methods do not include our um provided system meta prompt um but that will be coming
- [26:13](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1573s) preview: private preview will happen pretty soon after the business centrer launch event
- [26:33](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1593s) preview: the ETA for priate review is pretty soon so this 24x release so either 24.0 in April or 24.1 in May
- [26:53](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1613s) coming-later: the ETA for priate review is pretty soon so this 24x release so either 24.0 in April or 24.1 in May
- [27:14](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1634s) preview: public preview we will introduce a new process to satisfy responsible AI requirement
- [27:35](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1655s) subject-to-change: General availability ETA still to be determined or TBD depending on the preview feedbacks and result
- [27:35](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1655s) coming-later: the ETF for public preview will be the next release around 25th uh 25x version
- [29:17](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=1757s) preview: some of these cap abilities are considered previous specifications on Asia and AI side and in the toolkit we will not support any preview specification um at all
- [35:12](https://www.youtube.com/watch?v=ytYKvn2MiGs&t=2112s) preview: these are not production ready code but they just to show you how you can interact with the toolkit

Presenters (as heard): Derek Jew, Julia Sandy, Enrio, Derek, Julia, Sandy, Sarco.