April Aely spent more than 20 years leading technology initiatives at a Fortune 50 financial institution before leaving to build businesses of her own. Her latest venture is BloomVox, a marketing agency focused on creative, strategy, and visibility. From the start, she set out to change how an agency runs, how the work gets done, and what it’s like to be a client.
April is building both the agency and the technology behind it herself. Instead of a stack of disconnected tools, she’s creating one system where AI agents pass work to each other, support production, and keep clients up to date. She’s also getting the platform ready to license to other marketing agencies.
Each BloomVox client gets their own AI agent, called Bloom. Bloom knows the client’s brand, answers their questions, and takes their requests, then hands work to specialist agents behind the scenes. People lead the creative direction, make the strategic decisions, and approve everything before it’s published.
During her corporate career, April served as an Executive Director in technology and business systems, where she helped shape products, platforms, and complex operations. She also founded Peachity, which provides AI strategy, engineering, and fractional CTO and Chief AI Officer (CAIO) services to founders and businesses.
An engineer with an eye for good design, April loves making things and has strong opinions about how they should look, feel, and work. She’s especially interested in what AI now makes possible for independent founders, and how far experience, good judgment, and a willingness to question convention can take them.
At home, she’s mom to a fiercely independent daughter with a sharp wit and plenty of opinions of her own.
After two decades of building technology for someone else, April is building something that’s entirely hers. “I started BloomVox,” she says, “to create the kind of agency I’d trust with my own business.”
What is your typical day, and how do you make it productive?
I’m not sure there’s a typical day when you’re running two businesses and you’re also the person building most of the technology behind them. Some mornings I’m writing code or working through a technical problem. A few hours later, I might be designing a website, reviewing creative work, or figuring out how to introduce a new service to the market.
I spent more than 20 years in corporate environments where a good portion of the day could disappear into meetings about work that still needed to get done. I’m protective of my time now. I use AI extensively to help me build, research, and solve problems, but I’m still the one making the decisions.
I’ve also learned that being busy and being productive are two very different things. I can spend an entire day improving something nobody has asked for. Sometimes the most productive decision is to stop polishing it and move on.
How do you bring ideas to life?
I build them.
That probably sounds ridiculously simple, but it’s the biggest difference between how I work now and how I worked ten years ago.
I used to have ideas that would require a developer, designer, infrastructure, and a decent budget just to find out whether they were worth pursuing.
Now I can open my computer, sketch out an idea, build a working version, connect the necessary services, and start testing it myself.
I still do the research. I still think about the business model and whether anyone would pay for it. But I don’t spend months creating elaborate plans for something I could test in a few days.
The danger is that building has become so accessible that it’s easy to fall in love with making things.
So now the hard part is deciding which ideas are actually worth turning into a business.
What’s one trend that excites you?
AI agents that can coordinate and carry out real work.
We’ve spent the last few years watching AI generate text, images, and code. That’s impressive, but I’m much more interested in what happens when those capabilities become part of a functioning business.
At BloomVox, for example, every client gets their own AI agent called Bloom, who knows their brand and can help them manage the work we’re doing together.
A client can ask Bloom what’s happening with a campaign, request a change, or review something that needs approval. Bloom can coordinate the request, brief specialist AI agents, and move the work forward. Creative direction, strategic decisions, and publishing approvals still belong to people.
I think we’re going to see businesses operate very differently because of this. A small team, or even one founder, can now deliver what used to take a much larger company.
That’s an enormous opportunity for independent founders.
What is one habit that helps you be productive?
Questioning my own work.
I have a habit of looking at something I’ve built and asking myself whether I’d be impressed if someone else had created it.
Would I use this? Would I pay for it? Does the design hold up? Is the experience unnecessarily complicated?
I’m particularly critical of design and user experience. Something can technically work and still be a terrible product.
I also try to separate problems that genuinely need fixing from details I’m obsessing over simply because I built the thing.
I’d rather challenge my work before someone else has to, then get it in front of real people and find out what they think.
What advice would you give your younger self?
Start building your own things sooner.
I spent more than two decades building systems, solving problems, and contributing to the success of a very large company. I learned an enormous amount, and I don’t regret that experience.
But I wish I’d recognized sooner that I could take those same skills and create something of my own.
I also would have worried a lot less about following a conventional career path.
There are so many ways to build a career, a business, and a life. You don’t have to keep following the same path just because you’ve already spent years on it.
And I’d tell myself to buy the domain name.
Tell us something you believe almost nobody agrees with you.
I don’t think most businesses need 35 AI agents. Or 43. Or whatever number someone is celebrating on LinkedIn this week.
I keep seeing posts about people running their entire businesses with these massive teams of AI agents, and my first question is always: But why?
I’m building businesses with AI agents myself. I’m a huge believer in what they can do. But somewhere along the way, the number of agents became a bragging right.
Every additional agent introduces more coordination, more complexity, more opportunities for things to go wrong, and potentially more expense.
Sometimes a workflow needs several specialized agents. Sometimes one well-designed agent can handle the whole thing. And sometimes you don’t need an agent at all. A simple automation would do the job better.
I’d rather have five agents doing exceptional work than 40 agents with impressive job titles and questionable productivity.
Some of these companies are one efficiency audit away from massive AI layoffs.
The number of agents you have is not a business achievement. What is cool is what people are getting them to accomplish.
What is the one thing you repeatedly do and recommend everyone else do?
Use what you’re building.
I think it’s surprisingly easy to get so caught up in developing a product or service that you forget what the experience feels like to someone on the other side.
I’ve been putting my own businesses through the systems and workflows I’m creating. It’s one thing to design an onboarding process. It’s another to go through it yourself and discover where the experience becomes confusing, repetitive, or unnecessarily difficult.
You find problems that would never show up in a planning document.
The same applies to marketing. If you’re offering a service, look at how you explain it, how easy you make it to buy, and whether you’d find the offer compelling if someone else were selling it.
It’s a pretty effective way to discover where your own assumptions are getting in the way.
When you feel overwhelmed or unfocused, what do you do?
I try to figure out whether I’m overwhelmed by the amount of work or by the number of things competing for my attention.
Those aren’t always the same problem.
When you have a lot of ideas, it’s very easy to jump between them. You can make progress on six different things and somehow finish absolutely nothing.
I try to identify the one thing that would make the biggest difference if I finished it, and focus there.
Sometimes that means closing everything else and working through one problem. Sometimes it means getting away from the computer entirely.
What is one strategy that has helped you grow your business or advance in your career?
Learning to look beyond the immediate problem and understand the system creating it.
Throughout my corporate career, I worked with complicated business processes and technology platforms. What someone initially described as a software problem was often a workflow problem, an ownership problem, or a process that had become unnecessarily complicated over time.
The biggest improvements usually start with questioning how the work gets done. The technology comes after.
That way of thinking has shaped how I approach my own businesses today. Before I build anything, I want to understand the problem, why it exists, and whether the proposed solution is going to make someone’s life easier or just give them another thing to manage.
What is one failure in your career, how did you overcome it, and what lessons did you take away from it?
I think one of my biggest mistakes was waiting so long to take my own ideas seriously.
I had a successful corporate career. I held senior positions, worked on large technology initiatives, and had opportunities to solve complicated problems.
But I also had ideas of my own. And for a long time, they stayed ideas.
There was always a reason to put them off. I had a career, responsibilities, and a predictable income. Starting something from scratch was a much less certain proposition.
The mistake was thinking I’d eventually reach a point where the timing felt perfect and I had everything figured out. That point never came.
Eventually, I decided I’d rather find out what I could build than spend the rest of my working life wondering. I left my corporate career and started building businesses of my own.
The transition has been a big adjustment. Suddenly, you’re responsible for everything, including the parts of running a business that have nothing to do with your technical skills.
The biggest lesson has been that experience doesn’t eliminate uncertainty. You can be very good at your profession and still be a beginner at building a business.
You have to be willing to be one.
What is one business idea you’re willing to give away to our readers?
I’d build a follow-up and revenue recovery system for one specific type of service business.
Think contractors, specialty home service providers, or other businesses that spend money generating inquiries and preparing estimates.
A surprising amount of potential revenue can disappear after the initial conversation simply because nobody follows up consistently.
I’d build a small system that tracks outstanding estimates, identifies opportunities that have gone quiet, prepares personalized follow-ups, and shows the business owner what needs attention.
Start with one industry and one expensive problem. Don’t try to become another all-in-one business platform.
Charge a monthly subscription and prove the value through recovered opportunities.
It’s the kind of business that might not sound terribly exciting at a startup conference, which is probably part of its appeal.
What is one piece of software that helps you be productive? How do you use it?
Two, actually: Claude Code and OpenAI’s Codex.
I’m an equal-opportunity AI user. Whichever one is better for the task at hand gets the work. Between them, they’ve fundamentally changed how I work and how much I can build independently.
I use them to write and review code, investigate problems, work through architecture decisions, build interfaces, and connect the various systems that make my businesses run.
What I find remarkable is how much of the development process I can now handle myself. Something that once would have required several specialists can often be built and tested without assembling a team.
Of course, that doesn’t mean I trust everything these tools produce. Sometimes an AI coding agent will confidently propose a solution that’s completely wrong.
You still need to understand what you’re building, recognize when something isn’t right, and know when to push back.
What is the best $100 you recently spent?
Upgrading my Claude subscription to Max for a month so I could finish a software build.
That might not sound like a big decision, but I’ve built a lot using the lowest tier of everything in my workflow. I’m careful about what I spend, and it took me a while to talk myself into upgrading.
I was deep into development, making a lot of changes, testing features, and working through the inevitable problems that come with building software. I kept running into usage limits.
So I finally upgraded.
It wasn’t a particularly glamorous purchase, but it meant I could spend more time building and less time waiting.
When you’re building a business independently, anything that removes a bottleneck can be worth the money.
And considering how much time I wasted waiting for my five-hour limit to reset, I probably should have done it sooner.
Do you have a favorite book or podcast from which you’ve received much value?
I’m going to cheat a little and name a Substack newsletter instead: Ruben Hassid’s How to AI.
There is an unbelievable amount of AI content out there, and so much of it consists of people announcing that something is going to change the world without showing you how to do anything with it.
Ruben gets into the details. He tests tools, walks readers through his workflows, and doesn’t pretend to have all the answers.
Most of what I learn comes from building things myself and seeing what other builders have figured out. His newsletter is one of my best sources for the second part.
What’s a movie or series you recently enjoyed and why?
The IT Crowd.
I often have music playing while I work, but sometimes I’ll put on a show in the background. This British comedy about a chaotic IT department is one I keep coming back to.
It’s ridiculous, super awkward, and occasionally a little too familiar. Anyone who has spent time in enterprise technology will recognize more of it than they’d like to admit.
And after a day of arguing with AI agents and debugging software, watching someone else have a spectacularly dysfunctional day in IT is weirdly comforting.
Key learnings:
- Make sure someone wants it. AI has made it easy to build almost anything, which makes it just as easy to build the wrong thing. Find out who needs it and whether they’ll pay for it before spending weeks perfecting it.
- Give every AI agent a reason to exist. A few well-designed agents with clear responsibilities can accomplish an enormous amount. If you can’t explain why an agent exists, you probably don’t need it.
- Fix the process before adding technology. Some of the biggest improvements come from removing unnecessary steps and simplifying how work gets done.
- Use what you’re building. Go through your own onboarding, workflows, and sales process as if you were the customer. You’ll find what’s wrong before your customers do.
- Your experience comes with you. A long corporate career gives you skills, perspective, and judgment that carry into entrepreneurship. You’ll still be a beginner at running a business. Be OK with that.