Don’t Fear AI: Build Your Own Work Assistant This Week

Disclaimer: The views and opinions expressed in this article are those of the author and may not reflect the perspectives of IIBA.

This article is part of AI Wednesdays, an ongoing 2026 series that explores how business analysis professionals are using AI in real, practical ways. Written by a practitioner, each contribution shares experience-based insights to help you navigate AI with greater confidence—starting small, building familiarity over time, and applying AI where it adds real value to your work.

Artificial intelligence can often feel like a boogeyman in our industry, striking fear across many professionals who don't know how to incorporate it into their workflow, or worse, fear it will make their role redundant.

And that fear is real. But while the tech industry is facing many layoffs, we won't move forward in our roles by avoiding new technology altogether.

I’ll guide you as we go on this journey together. Instead of avoiding AI, I’ll help you create your own assistant. Along the way, I’ll reference some common AI terms (such as agent, instructions, and knowledge) that may be unfamiliar if you’re new to AI. To keep things simple, I’ll start with a few basic definitions.

• Agent: Artificial intelligence that can perform identified actions based on instructions and requests, utilizing reasoning and decision-making skills.
• Instructions: These inform the agent of what it needs to do. The instructions are similar to standard operating procedures (SOPs) but for AI.
• Knowledge: This is the information you provide to the agent that it can read through and reference. It isn't necessary to always have knowledge items, but they can be helpful and reduce the amount of detail you have to provide.
  
  

An Assistant to Help You

There’s so much that takes up your brain on a daily basis that isn't necessary and actually disrupts your flow state. It’s likely that at some point you’ve been working on writing user stories or acceptance criteria when suddenly you're called into a meeting. Now you need to jot down the notes afterward, capture your action items, and potentially be on additional calls about the ad hoc call that will switch your focus.

If you've read Deep Work by Cal Newport, you'll know that this constant context switching deprives you of your valuable flow state, which helps you to complete important tasks. Now you're back to writing your user story, and you have to remember what you needed to include. What if instead of being frustrated and trying to remember all the small items you needed to include, the assistant could capture notes and action items for you?

For example, you could have your assistant take a bulleted list of important items for the user story and craft it in the proper format. It could reference important information you've already documented, saving you from having to recall key details from many years ago.

By now, you can probably see where I'm going with this. Rather than fighting against AI or resisting it altogether, let’s instead engage with it and resolve some of the issues you deal with daily.

Assign Things to Your Assistant

I've outlined some things your assistant can help you with, but let's examine some concrete use cases that will help you implement them in your work.

1. Note-Taking

Most enterprise-level meeting tools have a built-in AI tool, like Zoom AI Summary, that people use for notes. However, I'm not a huge fan of the summary, and it takes some creative liberties at times.

A brilliant friend of mine recommended using transcripts instead, and it was one of the best things I've done. I use an agent that pulls key points and action items from the transcript and formats them the way I need. The agent also states what went well in the call and what could use some improvement.

I have found this incredibly valuable for saving time for my stakeholders and improving calls overall.

2. Writing User Stories

This can be really uncomfortable initially, because it might feel like you’re losing your job to AI. Remember, though, that you're the one providing information for it to generate user stories.

Unlike the note-taking agent, where you simply use the instructions to reference the transcripts, the user story agent can reference additional information (e.g., documents that list out test users, links to process maps, story formats, best practices from Agile or Waterfall). These can be files uploaded to the agent for it to reference or website links.

3. Policy Q&A

Depending on your industry, you might have to follow a lot of specific policies that can be difficult to recall.

For example, if you're in finance or anything that touches the government, there could be restrictive policies around legal or security issues. Capturing these within the agent can help you evaluate whether work you've done or stories you've created are missing information, or whether the strategy and process violate the company's policies.

4. Second Brain

It can be impossible to recall everything mentioned on calls, in meetings, process captures, and priorities. If your company uses Outlook, finding information in emails can be harder than it should be, even when the messages are only a few days old.

Depending on the tool your company uses, it could be connected to SharePoint, Notion, or other tools that allow you to reference notes captured. You won't have to expend as much brain power trying to remember all the different items when you can search via an agent. The agent can be trained using links to where you store your notes, can review your emails, and may even work with chat tools, depending on your company.

Best Practice Tips

When creating your assistant, keep in mind that AI is experimental. You're learning from AI as much as it’s learning from you. Here are some key things to remember, which may help simplify the process:

• Instructions should be hierarchical. The instructions at the top are the most important. While the agent should follow all of them, put the items that need to be followed at the very top of the list.
• If you're referencing documents, make sure they're focused on one topic. For example, I have my story-writing assistant reference a document that has profiles and people associated with those profiles to test with. I don’t add additional information about processes to that document because I want the AI to reference it and not have to read through documentation about processes.
• If you want to have things formatted, provide the AI with the format you want to reference. This can either be a document or written directly in the instructions, although I think a separate document works better.
• Be explicit in what you want the agent to do; don't try to have it write both notes and user stories. The agent should do one thing well.
• Iterate on what works and what doesn’t. Just because you created an agent doesn't mean you can't adjust the instructions and knowledge base it references.
• Always validate. The assistant is just that, something that helps you be more productive and alleviate stress. It isn't your brain. Everything it produces, whether you're asking chatbots a question or using an agent to generate work, must be evaluated. This is also why it can't replace you; someone has to review and correct its outputs if necessary.

Putting Your Assistant to Work

Now that you’ve seen a few key ways to leverage your new assistant, try creating a note-taking agent this week. It’s a simple way to lighten your workload and remove a task that’s pulling you in and out of flow.

Document what you want it to do and what information is needed, then write it up based on the tips outlined above. Once you have your agent or bot created, test the results and ensure it’s working as expected. If it isn't, tweak it by updating the knowledge, instructions, and anticipated results until it does.

The key to using artificial intelligence is to remember that it’s only as smart as the information you provide it. Guide it like you would an assistant new to your industry, and have patience in updating it until it understands it well enough to give a benefit.

AI isn't the boogeyman we should be afraid of. It’s more like a newcomer to the industry who requires a little bit of patience before they can start helping us achieve more.

Build confidence through proven practice. Explore the BABOK Guide to strengthen how you apply insight, judgment, and emerging tools like AI in your work.

---

About the Author