Why I left Argus at Southampton.

Argus could've been my company. I gave it away.

In April 2026, I had a clean shot at turning Argus into my pre-seed venture. The product was deployed across 150 users at 180 Degrees Consulting Southampton. The £3K Future Worlds Enterprise Prize had paid for the build. I'd pitched at Future Worlds. I'd recruited and trained a fifteen-person engineering team across four university societies that knew how to ship features.

The path was open. I could've taken the team with me, raised on the back of the deployment, and made Argus the venture I'd spend the next five years on. Most founders in my position would have.

I didn't. I left Argus at the university — codebase, team, institutional structure, all of it — and walked toward Substrate instead.

This is the essay about why.

What Argus actually was

Argus is a multi-agent AI product I built solo in 2024–25 to automate consultant workflows at 180DC Southampton. It started as my personal project and grew into something the chapter actually used. By the time I let go of it, fifteen contributors from four societies were shipping features, the Student Union was funding our LLM credits, and the consultancy itself had become Argus's anchor customer. There was a credible path to making it monetisable — selling into other 180DC chapters, productising the workflow agents, eventually charging for it.

I'm spelling that out because the move I'm about to describe — leaving it behind — only makes sense if you understand that the option to keep it was real.

Why most founders would've kept it

The default founder playbook says: when something works, double down. Argus worked. The technical core was de-risked, the customer was paying attention, the team was operational. Everything investors say they want to see was sitting there.

Most undergrads in my position would've turned it into their LinkedIn-headline venture. Some would've raised on it. A few would've gone to YC with it. The path was conventional and respectable.

The problem is I had a different bet to make.

The structural reason I left

Substrate is bigger than Argus could ever be. Argus is a workflow tool for a student consultancy. Substrate is the agentic platform for deep-tech engineering — chips, circuits, embedded systems, hardware. The TAM gap between them is roughly four orders of magnitude. The technical moat is incomparable. The customers, the investors, the timeline, the scale — all different categories.

Trying to run both simultaneously would've meant doing each at half intensity, which is the most reliable way to fail at both. Founders who can't choose are founders who don't compound.

So I had to pick. I picked Substrate.

But the question that took longer to answer wasn't which one — it was what to do with the one I wasn't picking.

Three options

Option one was to wind it down. Send a message to the team thanking them, archive the codebase, move on. Clean exit. But it kills the value the team had built and leaves the consultancy with a half-deprecated tool.

Option two was to keep Argus as a passive side project. Run it at 10% effort, do the bare minimum to keep it operational, make it a footnote on my CV. This is what most people do. It's also what most products in this state quietly die from — neglect dressed up as ongoing operation.

Option three was to leave it at the university properly. Hand the codebase, the team, and the institutional relationships over to the next generation of students. Structure the cross-society programme so it could recruit, ship, and operate without me. Let Argus become an institutional asset — something that lives at Southampton, that future undergrads can join, that the consultancy keeps using and improving. If the team eventually monetises it for the university, even better.

I picked option three.

Why option three is rare

Option three is uncommon for a reason. It requires you to do extra work for an outcome you don't directly benefit from. You have to write the documentation that lets the next person take over. You have to formalise the team's structure so it can absorb new recruits without you in the loop. You have to negotiate the institutional terms with the consultancy and the Student Union so the relationships transfer cleanly.

Most founders won't do this work. The math doesn't favour it: you're spending weeks of effort making a thing better that you've already decided not to own. The narrowly rational move is option one or option two.

But the narrow math is wrong, because it doesn't price in what the handoff actually accomplishes.

What it accomplishes

Leaving Argus properly does three things at once.

It honours the team I recruited. They came in on a pitch I made — that we'd build something real, that the programme would outlast me, that they'd have something concrete to point to. Sunsetting Argus the day I pivot would have made me a liar. Handing it over keeps the promise.

It builds an institution at Southampton. Most undergraduate ventures end the day the founder graduates. The cross-society engineering programme I structured was designed from the start to outlast me — to give future Southampton students a way to ship real products under real org structure, regardless of who's leading at any given moment. Whether the next cohort grows Argus, monetises it for the university, or pivots it into something better, the structure persists. That's worth more than any single product I could have walked away with.

It teaches me how to let go cleanly, which turns out to be the most important founder skill I'll need at Substrate. The default failure mode at any stage isn't picking the wrong bet — it's failing to put down the bets you already made. Companies fail more often from founders holding onto dead chapters than from founders missing live ones.

How this connects to Substrate

Substrate's whole thesis is platform compounding — building infrastructure that outlasts any individual agent, any individual customer, any individual engagement. Each Tracer modernisation feeds the next. Each agent inherits the trust of the last. The platform is bigger than any single product.

The version of me who builds that company is the same version who left Argus at Southampton. Both decisions come from the same instinct: care about the structure, not the spotlight. Build the thing that survives you. Trust the team you recruit to be more than your audience.

Argus is still alive. The team is still shipping. New recruits will join in September. The consultancy is still using it. None of that benefits me directly anymore.

All of it is exactly as it should be.

— Yassin