I asked an AI Agent to build a customized AI upskilling plan for a mid-size company. The result was super impressive.
I like to test new AI tools by asking them to do parts of my job. When I got access to Manus AI, a new, general purpose AI agent, I realized I could ask it to create an AI upskilling plan for on a mid-size company. I work with companies to help them upskill their employees for AI and increase AI adoption.
So asking AI to create upskilling plans is essentially outsourcing part of my work. But my job is also to figure out how to use new AI tools (like AI agents) in the workplace. While I might be inviting AI to do parts of my job, experimentation is part of understanding how AI works and what it can and can’t do. And spoiler alert, it did a very good job on this. It doesn’t feel like my job is at risk from AI, in part because so much of what I do is rooted in people engagement and sense-making about AI, two skills that AI won’t be replacing any time soon.
But back to the task at hand: This AI Agent did amazing work on researching best practices on AI upskilling for different teams and applying them to create a personalized employee upskilling plan for a mid-sized organization.
The result is mighty usable for any L&D or HR team who has ever been tasked with figuring out how to upskill employees for AI.
AI Agent vs. AI Assistant: What’s the difference?
First, a definition. We often use the terms AI assistant and AI agent interchangeably but they’re not the same.
Here’s the simplest way to explain it:
AI Assistant = Helpful sidekick
AI Agent = Autonomous teammate
Think of it like this:
An AI assistant is like a sharp intern. You give it a clear task (“Summarize this meeting”), and it gets it done right away but it waits for your instructions each time.
An AI agent is more like a proactive consultant. You give it a goal (“Create a company-wide AI upskilling plan”), and it figures out the steps, does research, makes decisions, and comes back with a full draft—even if you didn’t outline every step. This process takes a while.
Clarifying my goal: create a customized upskilling plan for generative AI
Like all AI tools, I started with a prompt. Prompts for AI agents don’t need to be super detailed. An AI agent prompt is focused on the goal. While I gave it some context, the AI agent asked me clarifying questions to ensure I got what I wanted.
My prompt focused on telling the AI agent I needed an upskilling plan for three different audiences in a mid-size company.
Manus asked me clarifying questions to ensure it understands what I want to do.
That’s all it took. From there the agent helped me break down the big, abstract question of “how do we upskill an entire company?” into something structured, practical, and doable.
The results
The report is long. 6,406 words long. Here’s the executive summary.
A personalized upskilling plan to improve AI adoption
While the report is too long to paste here, you can actually view the upskilling plan on the Manus platform. There’s a really cool feature that allows you to share their AI agent findings and see how it worked through the problem.
See the AI agent in action and read the final upskilling plan here
A few things stood out about using this AI agent:
It summarized the findings of the reports it found. In the screenshot above is a summary of multiple reports on executive education trends. Not something I asked for, but useful for future data points or narratives.
AI is great at structure. I knew the audience and goals, but the agent organized the content into logical sections, something that normally takes hours of brain juggling.
Speed + clarity = magic. It generated high-quality suggestions fast, and I could immediately spot where I needed to tweak things to match the company’s context.
It gave me ideas I hadn’t thought of. Like including a glossary of AI terms for non-technical employees or suggesting a change management plan. Those came from the agent, not me.
It felt less like outsourcing and more like working with a fast-thinking (and oddly confident) teammate.
How L&D teams can use AI in 2025
While I experimented with AI agents, you can do the same with other AI tools as well. I repeated the same task in ChatGPT using their Deep Research feature and got a solid upskilling plan (using the same prompt), including a section on how to do AI upskilling on a budget.
Fun fact: We offer both a train-the-trainer option for L&D and group learning with our AI learning labs
If you’re in L&D or HR, here are a few ways to use an AI agent or assistant to make an upskilling plan for AI
Draft your next learning strategy or proposal
Brainstorm training ideas across departments
Break down large goals (like “increase AI literacy”) into specific learning objectives
Compare training formats or approaches
Design rough timelines and rollout plans
You can try out Manus or use ChatGPT/Claude/Gemini:
Start with a real project with a goal. Don’t just “play”—ask it to help with something on your to-do list.
Give it clear roles. Tell it who it’s pretending to be (e.g., a workplace learning strategist).
Edit and refine. Treat the output like a rough draft with good bones.
Don’t overthink the prompt. Start small, then iterate.
Final thoughts on using AI for upskilling plans
I don’t believe AI will replace L&D professionals. We do so much more than make trainings. And I’m not convinced that people actually want to learn from AI avatars (who still have dead eyes), no matter how emotional their voices sound. L&D is evolving, just like all the other teams who are learning how to use AI in their roles.
Keep experimenting. And if you need some help, we’ve got AI learning labs for that.