Solving Hard Problems Starts Before Day One
Trial days can be a mixed bag. Maybe you’ve been there: you show up, shake some hands, admire the foosball table, get whisked off to the lunch spot (the one they’ll rave about daily, like it’s a company perk), and cap it off with a round of their go-to board game. It’s all very nice, very pleasant, and not particularly useful.
Mine wasn’t that. My trial day at Junction (formerly Vital) was structured, remote, and—importantly—paid. No gimmicks, no performative culture showcase. Just real work, with real people, in a setting that actually helped me decide if this was the right place for me. Here’s how it played out.
A brief word on me
I'm a Senior Backend Engineer at Junction, working on our lab testing product. If you’ve ever had a blood draw—at a hospital, in your home with a visiting phlebotomist, or through an at-home test kit—you know the testing phase. But what about scheduling the appointment? Rescheduling it? Making sure your sample actually gets to the lab? What happens if there’s an issue and someone needs to talk to the lab on your behalf? And what about payment or results analysis?
This broader remit is Junction's remit. Once I’m fully onboarded, it will be mine too. We're the custodians of keeping the system running smoothly at scale.
This is my second role in healthtech. Before this, I worked on food scoring, which is a different beast entirely. There, the challenge was about building the right model using the right data. Here, it’s about building lab testing fulfilment infrastructure on a national scale. It felt like a natural step up.
Why a trial day?
Back to the trial day. First, why a trial day? There are many reasons! From the candidate’s perspective:
- You experience a full day with the people you'll potentially be working with for years to come.
- You don’t just peek at the codebase, you live in it. You don’t theorise about their stack, you wrestle with it. By the end of the day, you’ve shipped something real.
- Depending on the task at hand, you might solve a problem from start to finish.
This all builds a lot of confidence. If you play your cards right, you might even deliver value before you've signed the contract. And don’t do thinking that this isn't a fair deal; it’s paid—remember that!
But this isn’t a one-way street—there's a host, too. Why would a company want to hold trial days?
- You see how a candidate works independently. How do they approach a real-world problem? How much initiative do they show? What fresh perspective do they bring? Compare this with the standard interview: an hour on a contrived problem with a solution that often feels rote. Don’t get me wrong—that has its place, but it tests something wholly different.
- You get real work out of it. Paying the candidate sets clear expectations. Sure, some might take advantage, but remember, this comes after the interview process when you've hopefully filtered out lazy or grifting candidates. The ones who make it this far want to demonstrate everything they've got in those eight hours.
I was intrigued. It felt fair.
On the day
I woke on a bright and chilly 3rd of January. 2025 was new news. By some miracle, my train from Edinburgh back to London the previous day ran on time without a hitch. My sleep was sound, my fridge fully stocked, the birds singing, and I had a 10 am call scheduled with Maitham, the CEO.
With a crack of my knuckles and a sip of tea (in reality, just tea), I sat at my desk. Time to get to work.
In that first meeting, Maitham was joined by Pires, both of whom I had met during my interviews. We exchanged best wishes for the new year as Maitham outlined the day: Pires would be my buddy and set my initial task. Pires and I spent the next hour or so going through the codebase. My task was to add a PATCH endpoint to an existing resource. Simple enough in concept, if not for a few crucial details:
- This was a live endpoint and my changes would go into production ("It's a Friday, so, you know, we'll hold off until Monday", which relieved a bit of pressure).
- The backend is a monorepo, and I needed to understand how the pieces within the blast radius of this endpoint fit together so I wouldn’t inadvertently start a fire before I had even officially begun (as that would certainly dash my hopes).
Once I had a decent understanding, I worked on the task independently for a few hours, pinging Pires every now and then and taking a quick lunch break.
At 2 pm, I had a call with Boris, a new face and the Head of Product. He outlined the lab testing product line to me, much like what I mentioned at the start of this post. The half-hour meeting flew by, and with more business context, I could see the pieces of the puzzle falling into place. But I was behind schedule: I had two tasks for the day, it was already mid-afternoon, and I hadn’t even raised my pull request yet. So, I threw all my attention back to coding my endpoint and managed to raise the PR before 3 pm.
My next and final task was to write an RFC on how to version this same resource. Updating the resource overwrites specific fields, some of which pertain to orders. It wouldn’t be acceptable for us to lose the exact details of, say, an order from last year simply because a linked resource was updated. This temporal aspect posed an interesting problem.
I proceeded to piece this puzzle together, drawing on the insights from my call with Boris, a few back-and-forth messages with Pires, and the (admittedly pretty awesome) Junction docs. At 4 pm, it was time for the company's bi-weekly virtual social, which I didn’t want to miss, so I knuckled down and focused on writing.
Happy Hour rolled around, and my RFC still wasn’t finished. I joined the social but excused myself early to complete my task. We played GeoGuessr. I was a solid geography student back in school and I'm fairly confident navigating new environments, but none of that mattered; GeoGuessr is hard!
At 6 pm, I had a wrap-up with Maitham and Pires. To be honest, I was shattered and offered my most unfiltered appraisal: long day, lots of context, interesting problem, big codebase, and a lingering question—why are the v3 endpoints in the v1 module (engineering folk will know the pains we go to to keep a codebase organised)? The relief after that 20-minute meeting was palpable. I popped the oven on knowing that, about 25 minutes later, I'd have a similar wrap-up with Beth, a meeting that went much the same way. I had a quick dinner and wind-down and, shortly after, fell into a sound, dreamless sleep.
My impressions
The next day, I carried a mix of excitement and relief. A few things really stood out:
- I was given a lot of autonomy and trust. Sure, my PR wasn’t merged on day one (I ended up merging it myself a few weeks later after I officially joined), and the task wasn’t rocket science—but it still held tangible value. I knew I had contributed, even in a small way, to improving the product.
- It was intense. Junction is a startup where moving fast is a given, but doing so with structure, clear deliverables, and dedicated time to focus led to a wealth of learning. Sometimes the most exhausting days are also the most fulfilling.
- Pires was there to support me right from the start. I never felt like I was left to fend for myself or compete for his attention. Prioritising people and ensuring everyone is brought up to speed is a big statement in itself, and it shows that, at Vital, people are the driving force.
Takeaways and outcomes
The trial day was integral in my decision to accept Junction's offer. When weighing up my options, I saw in Junction:
- A strong culture of continuous improvement: one that seeks feedback, supports each other, and is built on friendliness and high trust.
- Mature technical practices. After just a few hours with their main codebase, it was clear that a lot of thought had gone into designing their processes and workflows.
- A fast-paced environment that values autonomy, deep work, and active participation.
Transparency in benefits, incredible mission and strong compensation sealed the deal.
Now that I’m a few weeks in, I’ve noticed an unintended benefit. The effort I put into understanding the domain jump on my trial day jumpstarted my ability to absorb and retain knowledge during onboarding, making the whole process quicker and smoother.
Have you ever done a trial before joining a company? How did it influence your decision? And for those in hiring roles: have you considered introducing trial days? I encourage both candidates and hiring teams to give trial days more thought. Yes, there is a large up-front cost on both sides, but it reduces hiring uncertainty and the time invested can pay for itself in just a few weeks. I know it has for me.
Want to know if Junction is right for you? Do what I did. Try it and you’ll know. Check out our job board!