trajectory.

Near-term (rest of 2026)

Maintaining and scaling what's already in production. The voice agent at Multiskills IT crossed 10,000 calls last month and still needs daily tuning as the client's traffic grows. The real estate platform keeps adding features. The operations automations keep finding new corners to cover.

Shipping Vibe2Prod publicly. Both the CLI and the web platform are close to done. I want both out in the world before the end of the year, most likely making it open source to see how people use it, how it may be improved upon, and to have people judge the actual codebase.

Reviving Promptfluencer. I built it as an AI image and video generation platform, shipped it, got 20 users, and then shut it down because I lost discipline on scope and ended up drowning in features nobody asked for. Picking it back up with a much tighter focus this time around.

Looking at a Master's program in Canada this year. Deepening the formal engineering foundations while I continue building.

And landing the right full-time role, or working closely with someone deep enough in the space that the learning curve stays steep. I don't want to keep shipping client work indefinitely without being embedded somewhere that moves me forward.

Mid-term (the Master's, when it happens)

The program is coursework, not research. The goal isn't academic publication. It's broadening the engineering foundations I work from every day.

The practical side of building AI systems is mostly a solved path for me now. What I want is the parts around it: infrastructure, devops, cloud architecture, and a wider understanding of how AI actually lands inside different sectors. Healthcare, finance, logistics, government. The shape of the problem changes in each, and most of the interesting engineering happens in that shape, not in the model itself.

I also want the time to think clearly. Two years of building fast has been useful. Two years of building while also sitting in structured learning with people who think differently from me is a different kind of useful.

Long-term

I want to build something big enough, by the end of my 20s, to land on the lists that recognize people doing work that matters.

The category I care about is operational AI. Not chatbots, not copilots. Systems that actually run pieces of a business: scheduling, compliance, procurement, customer operations, back-office work that's currently held together by humans doing repetitive tasks that shouldn't be their job. There's a gap right now between the AI that demos well and the AI that runs reliably inside a company. Closing that gap is the most interesting engineering problem of the next decade, and most people talking about it haven't actually shipped one. I have, at small scale. I want to be working on it at large scale, whether as a founding engineer at a company doing it well or as a founder building one.

The honest version is I don't know which of those two paths I'm on yet. I'll know more once Vibe2Prod is in the world and I've seen how it may be adopted.