If you’ve spent any time on social media lately, you’ve probably been inundated with posts about AI and AI agents. There’s a lot of hype and, honestly, it can all be a bit overwhelming.
Let’s cut through the noise and explain what actually works today.
What is an AI Agent?
An AI agent is essentially AI with the ability to take action. Unlike traditional AI that simply responds to prompts, an agent can execute tasks, make decisions, and interact with systems autonomously.
Rather than asking AI to write an email and then copying it yourself, an agent could read the email, write a response, send it, follow up if needed, and update your CRM – all without your direct involvement.
Our First Internal Agent
We’ve built a number of AI projects for clients, but we’ve recently deployed our first internal AI agent, nicknamed Rakali.
Our Challenge
- Every code change needs review by a senior developer
- Keeps quality high but consumes valuable senior time
- Human reviewers can miss things despite their expertise
Our Solution
- Rakali now reviews every system change automatically
- Checks our growing list of quality controls
- Leaves detailed comments for developers to review
- Provides instant, consistent feedback
It’s early days, but Rakali is already providing a tonne of value. We’re still reviewing by hand in the meantime, but eventually senior developers will only step in for complex cases.
When to Use AI Agents
AI agents excel when you have:
- Repetitive tasks requiring judgement that can’t be reduced to simple “if this, then that” logic
- Multi-step processes where each step depends on the last one
- High-volume, skilled work that follows learnable patterns
When to Skip Them
Avoid agents for:
- Deterministic processes with clear rules
- One-off tasks that rarely repeat
- Processes where the cost of error is extremely high
- Tasks requiring instant responses (agents take longer to process than traditional code)
As a general rule, first see if traditional automation can get the job done. It’s still more reliable and cheaper to run.
What This Means for You
AI agents aren’t magic, but they’re the first practical step toward AI that works for you rather than with you. Start small, measure results, and gradually expand their responsibilities.
The businesses that figure this out first will have a significant operational advantage. The question isn’t whether AI agents will reshape how work gets done – it’s whether you’ll be early or late to the party.
See you next Friday!
