Your Guide to Microsoft AI: All You Need to Know for the PMO

Product & Tools
PMO & Governance
Peter Kestenholz
Founder - Projectum

When it comes to optimizing your PMO with artificial intelligence, Microsoft offers an array of options. Whether you need a chat assistant, a code generator, a custom model, or a large language model, Microsoft AI has something for you. This guide will break down the different types of Microsoft AI that are available and how they can help you achieve your goals.

Bing Chat: Accessible Search engine.

Bing Chat is an AI-enhanced web search that allows you to use GPT-4 and Dalle-E. You’ll recognize a lot of ChatGPT-3 in this, being able to ask questions, optimize prompts, and get answers. However, Bing Chat takes it a step further by being connected to the web and constantly learning.

It’s great if you just want to search the web or play around. However, Bing Chat isn’t suitable for sensitive data. Always weigh the risk factors before integrating it into your workflow – or consider Bing Chat Enterprise.

Microsoft Edge AI

Bing Chat Enterprise: A secure AI-web experience

Bing Chat Enterprise is a step up from Bing Chat. It’s bundled with your Office license and lets you chat with advanced language models like GPT-4, GPT-3, Codex, DALL-E, and Whisper

The great thing about Bing Chat Enterprise is that it doesn’t save or use your chat data to train its underlying models. This means that business data is completely protected and won’t leak outside the organization.

This also means you can use it for much more business-oriented tasks, such as

  • Understand the implications of a decision: “What are the pros and cons of offline marketing strategies?” and “How can I measure the success of my offline marketing campaign?”
  • Analyze data: “If we’re forecasting 7% EPS growth this coming quarter, how does our internal forecast compare with EPS growth in the top US public CPG companies?”
  • Summarize a work PDF open in Edge: “Recap the findings of this internal cybersecurity report and the top 3 concerns” and “What are best practices for addressing these kinds of security issues?”
  • Write better code faster: “Write a regular expression in Python that matches email addresses” or “How can I use the unit test module to write test units for my Python code.”

You can access Bing Chat Enterprise from and the Microsoft Edge sidebar using your work account. You need an eligible Microsoft 365 subscription to use Bing Chat Enterprise.

Microsoft 365 Copilot: Your Virtual Assistant for Outlook

Copilot is an AI assistant that can help you with many of your tasks in Microsoft 365, such as in Word, Excel, and Outlook. Copilot can understand your context and intent and provide relevant suggestions and actions.  For example, Copilot can help you:

  • Start emails quickly: Copilot can suggest email templates based on your recipients, subject line, and previous emails.
  • Generate a summary: Use Copilot to transform long email conversations into short summaries by extracting the main points, action items, and deadlines from your emails.
  • Plan an event: Copilot can help your team manage communications when preparing an event. Copilot can suggest meeting agendas, follow-up tasks, feedback surveys, and more.

Designed for office tasks, Copilot has a limited licensing pool of 300. It’s not just an add-on; it’s an investment with the potential to save countless hours, improve communication, and increase productivity.

Custom AI with Azure OpenAI

If you have the resources for a tailor-made solution, Azure OpenAI provides the toolkit for custom AI applications. You’re not bound by pre-made packages so you get a greater degree of freedom and control.

Azure OpenAI Service gives you the ability to customize and control your own AI models with OpenAI GPT-4, GPT-3, Codex, DALL-E, and Whisper models with the security and enterprise promise of Azure3.

Azure OpenAI Service lets you:

  • Fine-tune your models on your own data sets to achieve higher quality results than what you can get just from prompt design.
  • Scale from single to multiple GPUs for faster fine-tuning.
  • Leverage memory-efficient methods (such as INT4, LoRA fine-tuning) to reduce hardware costs by up to 90%.
  • Explore different fine-tuning methods and benchmark them to find the best performing model.

You can use Azure OpenAI Service with Python, REST APIs, or Azure OpenAI Studio. You need an Azure subscription and access granted to Azure OpenAI Service to use it. Finally, you need a solid data foundation to train your AI.

Do You Really Need to Make Your Own LLM?

With the myriad of solutions already available, creating your own Large Language Model (LLM) is often unnecessary. It’s a massive commitment and only recommended if you aim to be Microsoft-independent.

However, if you’re curious, xTuring is an open-source tool that lets you easily build, customize, and control your own personal LLMs4. xTuring allows you to:

  • Ingest data from different sources and preprocess them to a format LLMs can understand.
  • Scale from single to multiple GPUs for faster fine-tuning.
  • Leverage memory-efficient methods (such as INT4, and LoRA fine-tuning) to reduce hardware costs by up to 90%.
  • Explore different fine-tuning methods and benchmark them to find the best performing model.

You can use xTuring with Python or the command-line interface. You need a GitHub account and access to xTuring to use it.

Power PPM Logo blue rocket

With Microsoft, you get AI while staying in control

Microsoft’s AI solutions offer a robust framework, yet they give you the freedom to be in the driver’s seat. Choose the right tool for the job, but never forget: you’re still in control.

Deciding between these options isn’t a one-size-fits-all situation. Evaluate your PMO’s specific needs and consult the decision-makers. Whether you use Bing Chat Enterprise, Copilot, Azure OpenAI Service, or xTuring, you can always be in control of your data, models, and outcomes. Microsoft AI is designed to empower you, not replace you.

There’s also plenty of out of the box solutions that give you AI capabilities for the PMO, such as automatically generated risks or schedules, as well as secure custom chatbots based on your organizational data. Learn more about the AI capabilities of Power PPM here.

No items found.
No items found.
No items found.
No items found.
No items found.
Peter Kestenholz
Founder - Projectum

Peter Kestenholz is a successful entrepreneur and business leader with 20 years of experience from founding and growing the company Projectum. Peter is a recognized Microsoft MVP for 13 years straight, Fast Track Architect for the second year in a row the second year and a member of the Forbes Technology Council.

Read next...

show us where it hurts.

Get in touch with our brilliant team and share your current challenges and we'll share how we think we can help.
Get in touch