Make AI Work is a new newsletter brought to you by Future Work/Life, helping non-technical, AI-curious folks like you understand how AI can help you build great teams and careers. If you find it interesting, please share it!
Another week, another slew of AI developments. It's exciting, but difficult to stay on top of what matters for your business and career - let alone how to put these tools to use.
I've started this newsletter to share what you need to know and how it relates to your work. I promise to keep it brief, on-point and free of jargon (at least, without explaining what it means in plain English). I'm interviewing the people who are actually building and using tools like AI agents.
With that in mind, if you're working with agents or know anyone who would be interesting to chat to about making AI actually work, I'd love to hear, so get in touch.
Making AI agents accessible to everyone
Last summer, Dharmesh Shah, co-founder and CTO of HubSpot, recruited Sam Mallikarjunan to build what Shah envisions as a "professional network" for AI agents.
."Our entire vision for Agent.ai is to make it accessible for non-technical people to build, discover, customize, and use AI agents,"
In other words, while big tech companies are building AI tools that require teams of engineers and big budgets, Agent.ai is more like WordPress for AI - anyone can use it to build something useful.
What makes something an AI Agent?
As Sam explains:
“I define an agent as a piece of software that can take an action autonomously on your behalf.”
Think of an AI agent like a personal assistant who can do specific tasks for you automatically. Just as you might ask an assistant to 'prepare my meeting briefs every morning' or 'keep track of my expenses', an AI agent can be set up to do these jobs without you having to ask each time.
The difference between this and regular automation?
An AI agent can make simple decisions and handle things that aren't exactly the same every time - like figuring out which parts of someone's LinkedIn profile are most relevant for your upcoming meeting.
But here's the key insight: you don't need to be technical to build one.
"All you need to do is be comfortable with the workflow building tools that you'd have with email automation or social media scheduling."
From personal projects to work tools
When I asked Sam about his first experiments with Agent.ai, he didn't jump straight to business processes or productivity tools. Instead, his favourite project was something quite different:
“I arguably find that not as interesting as my Hallmark movie viewing list generator.”
And he's onto something - starting with a personal project often leads to the best learning.
So, here are three examples of what's possible, from the playful to the practical (all agents Sam’s built himself):
1. The Hallmark Movie Curator
A creative example of personalisation:
Processes YouTube transcripts and summaries
Creates Hallmark movie viewing recommendations for Sam
This is a great example of how agents can organise content in ways major platforms haven't attempted
This might sound frivolous, but it solves a real problem - finding the right content in a sea of options. Instead of scrolling endlessly through recommendations that don't match your taste, the agent learns what you like and finds matches. The same approach could help you filter industry news, research papers, or any other content you need to stay on top of.
2. The Chess Coach Assistant
A personal learning tool that:
Fetches YouTube chess tutorial transcripts
Reviews the content
Creates practice questions
Tests Sam’s understanding before sessions with his chess coach
Think of this like having a study buddy who watches all your chess tutorials for you, picks out the key points, and quizzes you before your actual coaching session. Instead of showing up unprepared or spending hours reviewing videos, the agent helps you make the most of your expensive coaching time.
3. The Meeting Prep Agent
Sam uses an agent that every morning:
Scans his Google Calendar
Identifies who he's meeting
Uses Clearbit to enrich data about attendees
Fetches their recent LinkedIn posts
Creates a briefing of relevant information about each person and shares it with Sam pre-call
Instead of spending 15 minutes before each meeting frantically Googling who you're about to talk to, this agent does all that research for you automatically. It's like having a really efficient personal assistant who makes sure you never walk into a meeting unprepared
The key thing about all these examples? They're not replacing humans - they're doing the time-consuming prep work that lets humans focus on what matters: having better meetings, learning more effectively, or simply enjoying their free time.
What you can do today
Start simple: Build something personal first - like Sam's Hallmark movie curator
Think customisation: Focus on a very specific need rather than trying to build one agent that does everything
Join the community: Learn from the thousands of agents already built on the platform
Thinking ahead, what’s one AI-powered tool that, built properly, would change your work?
And would you trust an AI agent to handle critical work for you? Why or why not?
I hope you enjoyed this first edition of Make AI Work. I’d love to hear your feedback so do reply and let me know.
Have a great weekend,
Ollie