Category: AI

Exam AB-210: Dynamics 365 Sales AI Consultant Associate

Indeed the 3rd exam related post in just over a week – it’s a busy (new) certification release season at the moment!

This time it’s the new AB-210 exam, focusing on Dynamics 365 Sales and AI (of course!). It’s nice to see that there’s a dedicated Dynamics 365 Sales exam back now – most of us will remember the MB-210 exam that was around for a number of years, but which was retired at the end of November 2024. What happened was that a new exam at the time (the MB-280) was released, which rolled together Dynamics 365 Sales with Dynamics 365 Customer Insights.

I never fully officially understood the reason for this, given that the roles in reality are quite different, and did comment at the time (MB-280: Microsoft Dynamics 365 Customer Experience Analyst) that I wondered how well it would stand the progress of time.

AI and sales capabilities seem generally to go well together – Microsoft has publicly demoed at large conferences the Sales Agent multiple times, showing how it can help qualify leads, and handle engagments with customers. To be honest I quite like this in general, though for implementation I do keep my (slightly skeptical) eye on it, to ensure it’s working in the right way.

The official description of the proposed exam candidate is:

As a candidate for this Microsoft Certification, you design and configure AI-enhanced sales solutions by using Dynamics 365 Sales, Copilot in Dynamics 365 Sales, and agent capabilities to help sellers work more efficiently throughout the lead-to-cash process. You translate business requirements into practical seller workflows enhanced with conversational intelligence, predictive insights, guided automation, and secure data access.

In this role, you work closely with sales, operations, and IT stakeholders to help ensure that solutions align with revenue goals and process optimization.

You perform the following design and implementation tasks:

  • Configure Dynamics 365 Sales core features.
  • Deploy, manage, and monitor agents in Sales.
  • Implement collaboration features.
  • Tailor AI-powered intelligence features.

It is highly recommended that candidates complete training in intermediate-level Microsoft Power Platform configuration before taking this certification exam. Additionally, you must have functional knowledge of:

  • Building Power Automate cloud flows.
  • Interpreting an organization’s sales processes and seller experience.
  • Building and extending model-driven apps.

The overall information for the exam can be found at Microsoft Certified: Dynamics 365 Sales AI Consultant Associate, and there is an official Learning Path available for it.

I do like that the exam content overview calls out that Power Platform knowledge & configuration is highly recommended. Obviously Dynamics 365 is built on top of Power Platform, and having this knowledge (ie the ability to customise & extend with Power Platform capabilities) is key to well thought through implementations.

As I’ve posted before around my exam experiences, it’s not permitted to share any of the exam questions. This is in the rules/acceptance for taking the exam. I’ve therefore put an overview of the sorts of questions that came up during my exam. (Note: exams are composed from question banks, so there could be many things that weren’t included in my exam, but could be included for someone else!). It’s also in beta at the moment, which means that things can obviously change for when it comes out of beta.

I’ve tried to group things as best together as I feel (in my recollection), to make it easier to revise.

  • Setup & Data
    • Environment creation & provisioning
    • Document management options & requirements
    • Enabling AI capabilities (Copilot, Sales Agent etc)
    • Configuring & customising forms
    • Configuring & customising views
  • Outbound calling
    • Configuration
    • Security requirements
  • AI Capabilities
    • Getting access to AI capabilities for users (deployment, security etc)
    • What the different AI agents & modes are, when to use them, and the behaviour of each
    • What blueprints are, how to use them, how to modify them
    • How AI agents handle communication re-tries
    • Creating custom agents
    • Analysing AI agent behaviour (runs, outcomes, metrics etc), monitoring information
    • Using AI to summarise records & ask for information
    • Ways to handle AI usage billing (what options are available, where to do this, how to do this)
  • Leads & Opportunities
    • Setting up & configuring predictive lead scoring models, requirements for implementing this
    • Understanding lead to opportunity conversion process, and continuing through to a final sale
    • Understanding sales goals, configuring sales goal/metrics/KPI’s, configuring rollup queries for aggregation
    • Assignment behaviour for leads to users, how this works, configuration for this
  • Products
    • The different ways to handle products (eg units, bundles, price lists, product families)
    • When each one should be used, and requirements for them
    • How to use the different components to configure specific scenarios
    • Relating products together
  • Pricing
    • Different ways to approach pricing products (eg singly, as a bundle, etc)
    • Handling multiple territories
    • Handling multiple currencies
    • Configuring price lists
    • Handling expired price lists & system behaviour
    • Handling discounting
  • Mobile app
    • Setup & configuration
    • Data synchronisation
    • Security setup & requirements
    • Push notifications
  • Power Automate
    • Understanding when to use different trigger types (automated/manual/schedule)
    • Usage for scenarios requiring approvals
  • Business process flows
    • What they are, and what they should be used for
    • How to configure, moving between stages, understanding how they work

I hope that this is helpful for anyone who’s thinking of taking it – good luck, and please do drop a comment below to let me know how you found it! I’d also be interested in your thoughts/opinions around the direction that Microsoft has taken for this!

Exam AB-620: Design and build integrated AI agent solutions in Copilot Studio

We seem to be on a roll here over the last month or so with new exams being released (& its not over yet!). With all of the emphasis on AI & agents, I decided to go take the new Copilot Studio exam to see what it would be like.

Given that I have a decently passing familiarity with Copilot Studio (as I use it for projects, and actually do get hands on with it quite a bit of the time), I felt that I’d be in a good place to handle it without any revision. Obviously this could have been a bold move, and it’s up to everyone to make their own decisions about how much to revise (or not revise)!.

Copilot Studio has moved on from when it first came onto the scene (and for those who remember, it used to be called Power Virtual Agent, or PVA). Nowadays it supports coding within it, but it also can serve as the front end for other Microsoft AI capabilities, such as Microsoft Foundry models.

This is also the first time that it’s been featured for its own exam – previously it got rolled into other exams (such as the PL-100, PL-200, etc), where it was just one of the components being covered (and covered in a lightweight manner, at that). With the focus from Microsoft now heavily on it though, it’s now taken a step forward into the spotlight by itself.

The official description of the proposed exam candidate is:

As a candidate for this Microsoft Certification, you’re a professional developer or advanced builder who builds, extends, and integrates custom agents for enterprise-grade solutions. You typically work as an IT application developer, consultant, or independent software vendor (ISV) partner focused on creating scalable AI solutions for organizations or customers.

For this exam, you should be familiar with Power Fx, Microsoft Dataverse, Microsoft Power Platform environments and components, Microsoft 365 Copilot, Microsoft Foundry, and adaptive cards.

You need intermediate knowledge of generative AI concepts, including models, orchestration, retrieval-augmented generation (RAG), Model Context Protocol (MCP), Agent2Agent (A2A) protocol, and more. You should also have experience with prompt engineering and with REST APIs and integration patterns. Additionally, you need experience configuring agents with basic knowledge sources, instructions, tools, and topics in Microsoft Copilot Studio.

As a developer who works in Copilot Studio, you:

  • Integrate agents with Microsoft Foundry.
  • Integrate agents with Model Context Protocol (MCP) servers.
  • Integrate agents with custom connectors.
  • Integrate agents with APIs.
  • Integrate agents with Microsoft Fabric.
  • Automate tasks with computer use.
  • Integrate agents with connectors.

You create:

  • Multi-agent solutions.
  • Agents with enterprise knowledge sources (such as ServiceNow, SAP, and others).
  • Advanced agent topics and tools.
  • Computer-using agents.
  • Agents that perform advanced actions via APIs.

You collaborate with Microsoft 365 administrators, Microsoft Power Platform administrators, Microsoft Copilot administrators, Copilot Studio agent builders, Copilot Studio administrators, Foundry administrators, agentic AI business solutions architects, and Copilot Studio architects.

The overall information for the exam can be found at Microsoft Certified: AI Agent Builder Associate, and there is an official Learning Path available for it.

As I’ve posted before around my exam experiences, it’s not permitted to share any of the exam questions. This is in the rules/acceptance for taking the exam. I’ve therefore put an overview of the sorts of questions that came up during my exam. (Note: exams are composed from question banks, so there could be many things that weren’t included in my exam, but could be included for someone else!). It’s also in beta at the moment, which means that things can obviously change for when it comes out of beta.

I’ll freely admit that there was a LOT more focus on MCP capabilities than I had expected there to be, but I guess that again this is natural, given how Microsoft is moving at the moment.

I’ve tried to group things as best together as I feel (in my recollection), to make it easier to revise.

  • Copilot Studio
    • Component/node types. What they are, how/when to use them
    • Using topic variables
    • Timeouts
    • Concurrency
    • Sensitive data & Using type ‘secret’ – what this does and why to use
    • Generative answers – how they work, limitations, what to know, how to configure & ground them
    • Computer Use
    • Connecting with Microsoft Graph
    • Connecting to other agents – how to do this, how to configure, what to use
  • Connector types
    • Standard connectors (ie connectors provided by Copilot Studio). When to use them, limitations
    • Custom connectors – what these are, why you’d use them
  • Security
    • Authentication types (API, OAuth 2)
    • Query delegation
    • DLP policies
  • MCP servers
    • What they are
    • Connecting to them
    • Security with MCP servers
    • Authentication types
    • Usage of AI with MCP servers
  • Azure AI Search
    • Connecting to knowledge index
    • Configurations
    • Security
  • Solution Types
    • Default vs Unmanaged vs Managed
    • Environment variables
    • Creating solution
  • Application Lifecycle Management (ALM)
    • What this is, and why it’s needed
    • What approaches can be used, why to use them
    • What’s needed to set up ALM
  • Monitoring & Troubleshooting
    • Reporting on deployed agents
    • Evaluating usage of deployed agents
    • Identifying issues & errors
    • Stopping runs

I hope that this is helpful for anyone who’s thinking of taking it – good luck, and please do drop a comment below to let me know how you found it! I’d also be interested in your thoughts/opinions around the direction that Microsoft has taken for this!

Exam AI-901: Microsoft Azure AI Fundamentals

With a massive amount of focus on AI across the Microsoft platform, I decided to sit the new AI-901 exam, which is the new Azure fundamentals exam. I’m far from being an Azure architect, but will freely admit a decent amount of familiarity with a lot of Azure components, especially the AI stuff. Having previously passed the AI-900 a while back, I was expecting the exam to be up to date with technical developments, but wasn’t FULLY prepared for what it was actually like…

Now obviously all Microsoft AI capabilities, regardless of where they’re surfaced through, actually sit (somewhere) within Azure. After all, Azure is the Microsoft cloud platform itself (well, until someone decides to rename it, of course).

My expectations for going into the exam (with admittedly very minimal preparation for it) was to cover the basics for AI within Azure, similar to the way that the AI-900 exam was. Whilst this was somewhat the case, it didn’t necessarily stay within the bounds of my expectations.

The official description of the proposed exam candidate is:

This certification is intended for individuals who want to start working with AI solutions built on Azure. It is suitable for learners from technical backgrounds, including aspiring junior developers who are starting to incorporate AI capabilities into applications. As a candidate for this certification, you should have familiarity with the self-paced or instructor-led learning material.

This certification assesses your ability to show the conceptual knowledge and practical understanding needed to work with AI solutions on Azure, including:

  • Understanding core cloud concepts, such as services and resource deployments
  • Using Microsoft Foundry to deploy models and implement single-agent solutions
  • Recognizing how client applications are put together and how AI models and services are consumed within those solutions
  • Understanding Python code examples that call AI models and services

This certification is intended to validate skills commonly used when performing tasks such as:

  • Adding AI workloads, including language, vision, and generative AI, to software or IT solutions
  • Exploring and using AI features in applications as a junior or entry level developer

The overall information for the exam can be found at Microsoft Certified: Microsoft Azure AI Fundamentals, and there is an official Learning Path available for it.

As I’ve posted before around my exam experiences, it’s not permitted to share any of the exam questions. This is in the rules/acceptance for taking the exam. I’ve therefore put an overview of the sorts of questions that came up during my exam. (Note: exams are composed from question banks, so there could be many things that weren’t included in my exam, but could be included for someone else!). It’s also in beta at the moment, which means that things can obviously change for when it comes out of beta.

My main shock was the number of questions on Python code, including needing to select the right code syntax to use. Whilst I do understand that Microsoft is aiming to make Fundamental level exams/certifications more ‘technical’, I do feel that this is much more technical than the audience should be experiencing. I’ve also fed this back as feedback into Microsoft.

I’ve tried to group things as best together as I feel (in my recollection), to make it easier to revise.

  • Analysis
    • Analyser types (audio, document, image, video). What each type is, how to configure them, and when to use them
    • Defining schemas for data extraction
    • How to extract content for analysis
  • Python
    • Using the Python SDK
    • Python code syntax and commands
  • Microsoft Foundry/Foundry Models
    • How AI models actually work when using/interfacing with them. Behaviour, access to content, prediction etc
    • LLM evaluations – comparing costs and capabilities
    • Creating, configuring, deploying, updating
    • Model temperature, inference
    • Minimising model bias, ensuring fairness
    • Connecting to a deployed model
    • Message structures for Foundry projects
    • Agent Evaluators – what they are, how to use them
    • Using Azure Content Understanding
  • Usage for models
    • Using Azure functions
    • Encoding images – data types
    • Voice Live (audio to text)
    • Azure speech SDK, and classes to use
  • Prompts:
    • Agent prompts. What are they, how are they used, why you should use them
    • System prompts. What are they, how are they used, why you should use them
  • Microsoft Responsible AI Principles – what they are, what are example of them
  • Why humans are still important to be involved in processes

I hope that this is helpful for anyone who’s thinking of taking it – good luck, and please do drop a comment below to let me know how you found it! I’d also be interested in your thoughts/opinions around the direction that Microsoft has taken for this!

AI, Microsoft Copilot, and Copyright

Firstly, a Happy New Year to everyone – I’m sure that you have amazing plans for 2024, and wish you the best of luck with them!

Over the last few months, I’ve been keeping a close eye on Copilot (since it was announced as being in GA), and various happenings around the wider AI scene. It seems almost impossible to find someone who is NOT aware of the OpenAI happenings towards the end of 2023, and so many more conversations now include mention of ChatGPT, Azure OpenAI, etc.

But there’s one item I’d like to pick up on & discuss, which given the news events of last week, is extremely pertinent. This is the topic of copyright, which is a very important topic to understand. However in order to understand it properly, we need to understand how AI offerings are generated in the first instance.

All of the various AI offerings rely on LLM’s. This stands for ‘Large Language Model’, and these are deep learning models that are pre-trained on vast amounts of data, and by vast, the number of data points are truly staggering:

Incidentally, users should consider the use case that they’re using AI for, and look to use the best optimum model for it. As an example – if wanting to find out the best routine for making a cup of coffee, it is unlikely that a GPT-4 model would be needed – a suitable model could be one with less parameters in it. This is in part due to the amount of resources needing to be used to run queries on more advanced data models.

When using an AI offering, these datasets are used to present answers back to the users. Some AI models are not limited to just their LLM datasets, and are able to actively trawl & access websites as well for more up to date information.

But there are two inherent problems with the way that this can work:

Data

When users interact with chatbots or other AI capabilities, they’re inputting data into them. This data could be used by the AI capability to further train models forward, ingesting the data provided to them. Given that data could be sensitive or proprietary, this can be problematic. Not all AI organisations use data just for processing, as has been discovered by various users.

Copyright

Given that responses back to users (whether in text format, image format, or other formats) are based on the datasets from the underlying LLM’s, users could potentially be provided with information that is actually copyright, and which they’re not able to use. This is very problematic, as it can result in users passing off material as their own, whilst it actually belongs to someone else.

Microsoft’s approach

Microsoft’s approach to AI capabilities has been made extremely clear. You may be familiar with seeing slides similar to the following slide:

What this means is the following:

  • Microsoft will not use any customer data to train AI models & capabilities for any standard AI offering
  • Any custom Copilots created by Microsoft customers will remain their own – Microsoft will not use data or capabilities from customer created collateral within the Microsoft AI offerings. This means that a bespoke Copilot will only offer its functionality to the customer that created it – other organisations, even within the same sector & creating Copilots themselves, will not benefit from this

Microsoft has also confirmed publicly that any information generated through the usage of Copilot or Azure OpenAI can be used without concern about copyright claims (Microsoft announces new Copilot Copyright Commitment for customers – Microsoft On the Issues). In fact, Microsoft has even gone so far as to say that Microsoft will assume responsibility for any potential legal risks involved.

If a third party sues a commercial customer for copyright infringement for using Microsoft’s Copilots or the output they generate, we will defend the customer and pay the amount of any adverse judgments or settlements that result from the lawsuit, as long as the customer used the guardrails and content filters,

Brad Smith, Microsoft Chief Legal Officer

This is quite important – customers are able to use Copilot & Azure OpenAI capabilities, and be assured that they will not have to be concerned about copyright issues or challenges. There are of course some conditions around this, in the way that prompts & interactions need to be handled (see Customer Copyright Commitment Required Mitigations | Microsoft Learn for further information on this).

Microsoft has this called out specifically within their Universal License Terms, available to view in full at Microsoft Product Terms.

Recent news events

With the announcement last week that the New York Times has filed suit against the OpenAI Corporation & Microsoft, this is very timely to look at (New York Times Sues OpenAI and Microsoft Over Use of Copyrighted Work – The New York Times (nytimes.com)).

The implications of such a lawsuit will affect how AI capabilities will be able to be created & used on an on-going basis. Copyright is of course very important to respect, and it will be quite interesting to see how this plays out. Having taken a look at some of the material included in the lawsuit, there are most definitely similarities between the New York Times information, and the AI generated output.

So, in my opinion, this is going to be a very interesting space moving forward, and I look forward to seeing how it goes, and any effects that it has on the usage of AI within organisations.