An Open Letter to MailerLite: What Email Platforms Need To Change for the Age of Human-Agent Collaboration

To the MailerLite team,

I’m building a product that, amongst other things, runs email operations on behalf of small businesses. My users are the kind of solo founders and small teams who make up a large part of your customer base. Many of them already use and love MailerLite. Thus, I chose to build on you to enable them to continue to benefit from the great work you have put into building your product and serving your users. However, I have found that the API you offer stops short of what AI agents working on a customer’s behalf now need. This letter is written in a spirit of collaboration, not as a complaint, but with the hope that what I share could help to make you the platform that leads this shift rather than gets worked around.

I’m sure you are well aware of the unique point in history we meet as software builders. For years, APIs were built for one of two callers: a human clicking in a dashboard, or a rigid integration moving data between systems. A third caller has arrived: an AI agent working alongside a human, doing the parts that are repetitive and handing back the parts that need judgment. That collaboration only works if the platform exposes the same capabilities to the agent that it exposes to the person in the dashboard, and in the same way. Where it doesn’t, human-agent collaboration is impeded.

Here is where I hit a wall:

1. The email designs and the API live in separate worlds.
When an agent creates a campaign as HTML, the customer can no longer edit it in your drag-and-drop UI, it becomes a block they can’t touch in the tool they know. And the reverse fails too: an email the customer built in your editor can’t be reliably read back out as editable design. The round-trip is broken at both ends.

The fix: let the API and the visual editor share one format, so designs can pass cleanly between agent and customer in both directions.

2. Agents can’t do what the dashboard can.
Automations are the clearest example of this: the API lets me create an empty, name-only draft, but there’s no way to configure a trigger or a step programmatically. So when a customer’s signup flow needs building or adjusting, software can’t do it for them, even with full permission, and the work falls back on the human.

The fix: bring the API to parity with the Editor/Dashboard, so anything a customer can do by clicking, an authorized agent can do too.

3. Basic building blocks that an agent could use are missing.
Some of the most ordinary operations simply aren’t there. For example, I can’t duplicate a campaign. Copying last month’s newsletter to use it as a starting point for the next edition, a common task for anyone running a newsletter, has no endpoint and must be done by hand. Nor can I send one email to one person in response to something they did; the community workaround of adding them to a group and letting it send isn’t reliable, and while I know MailerSend exists for this, its absence from MailerLite itself feels like a real gap, especially in the age of agents.

The fix: provide these building blocks directly, so agents don’t have to lean on workarounds that break under real use.

None of these are unreasonable design choices for the world they were built for. The point is that the world changed. The unifying principle of my suggested fixes is this: anything a human can do in the dashboard, an authorised agent should be able to do through the API. But there is an opportunity to go even beyond parity. Agents can do things no human reasonably would, things that would have been taxing or complex, and the MailerLite team has the opportunity to build for what agents make newly possible.

I’d genuinely rather build deeper on MailerLite than route around it, that’s why I wrote this open letter rather than quietly moving on. I truly believe that this is an exciting time to be building software. Contrary to those who predict a ‘Saaspocalypse’, I believe there is incredible opportunity for those who seize the opportunity to help agents better serve people. The platforms that address gaps like the ones I have identified will succeed because they will be the ones that AI-powered tools, platforms and builders recommend to the next wave of small businesses. I hope that you choose to be among them.

With my best wishes,

Esther Kuforiji