CreateLixir is a small team building a large idea. This is how we work, how we make decisions, and how we treat each other — described the way we actually behave, not the way a poster would say we do.
A company's culture is the sum of the choices people make when the founder isn't in the room. Every product decision, every interaction, every commit is a small vote for the kind of place we're building — and, over time, the kind of product we get to build.
We take this seriously because it compounds. A team with a strong culture ships better software faster and hurts each other less doing it. A team without one relearns the same arguments every quarter.
These aren't slogans — they're the answers to 'how do we usually handle this?' when the answer needs to be legible to a new teammate.
We make decisions that will still make sense years from now.
Quarterly optics matter less to us than what the platform looks like on its fifth birthday. When two options tie in the moment, the one with the better long tail wins — always.
Everyone is trusted to see problems through.
If you notice something is off, you're empowered to fix it — not to file a ticket about it. Titles describe past work; ownership describes present intent. We optimise for the latter.
Attention to detail is not optional.
Software gets shaped by a thousand small decisions. We treat each of them as if the reader in six months will notice — because they will. Care compounds.
We stay curious about our own opinions.
The state of the art moves weekly. We keep experimenting, keep reading, keep asking colleagues to explain things we already think we understand. The moment we stop being curious is the moment we stop being useful.
Clear thinking, simple solutions.
The hardest engineering problem is usually explaining what you built to someone who wasn't in the room. If it's hard to explain, it's probably not finished. We aim to leave things simpler than we found them.
Ideas matter more than titles.
The best product decisions come from tight loops between designers, engineers, and the people using the product. We move disciplines through each other's work early and often, and we treat 'that's not my area' as a signal to lean in, not to retreat.
Constructive feedback, welcomed perspectives, no punchdowns.
The bar for how we treat each other is higher than the bar for the work. Every teammate deserves to be able to say a hard thing — and to have it heard on the substance, not the tone.
Every release is an opportunity to learn.
We ship, we watch, we listen, we adjust. Post-mortems aren't punishments — they're the mechanism by which the platform gets steadily better without anyone having to be a hero.
Regardless of title or discipline, everyone at CreateLixir is here to help developers, founders, and teams build software more effectively with AI. That's it. That's the job.
A designer's mock, an engineer's commit, a support reply, a marketing post — every artefact we produce succeeds or fails on the same test: does it help someone build better software? When the answer is no, it doesn't matter how elegant the artefact is; it isn't done.
Culture shapes decisions at the whiteboard. These principles shape decisions at the keyboard.
The seams belong in the right places.
Every module has a job and knows its edges. We invest early in the boundaries between subsystems because that's the tax that never gets cheaper to pay later.
Designed for the load we plan to have.
We don't build for imaginary web-scale. We build so that when a real customer's usage grows tenfold, nothing needs a rewrite — just a raised limit.
Boring is a compliment.
Users don't want thrilling infrastructure. They want the platform to be there tomorrow, and Wednesday, and next February. We instrument, we alert, we practise.
Speed is a design decision.
Latency is a feature. Every surface gets performance budgets during design — not tuned after the fact when the shape of the code has already committed us to a slow path.
Built to be used by everyone.
Keyboard navigation, semantic markup, colour contrast, ARIA where it matters. If a surface isn't accessible, it isn't shipped — and this is not a nice-to-have with a deferred date.
Defensible by default.
Secrets stay out of prompts. Tokens are scoped. Trust boundaries are drawn deliberately. We treat every piece of user data as something we've been entrusted with, not just stored.
Written to be read six months later.
Code is written many times and read many more. We favour the version of the code a rested colleague can understand at 3pm — not the version that showed the most cleverness at 3am.
Every surface is a decision, not a default.
Nothing ships because we ran out of time to design it. If a surface is generic, that's an intentional choice we can defend — otherwise it deserves the same care as the flagship features.
Being distributed isn't an accommodation — it's the shape of the team. Everything below flows from taking that seriously instead of pretending it's temporary.
Great remote work isn't videoconference-driven office work with worse lighting. It's a different discipline: high-quality writing, calmer decision-making, and a shared bar for what "communicated" actually means.
Most decisions happen in writing so anyone can join later, catch up quickly, and disagree in-line.
If a decision isn't written down, we treat it as if it hasn't happened. Real-time meetings produce written outcomes.
We design our rituals so no one has to be online at 2am for their contribution to count.
Roadmap, decisions, and reasoning are visible to the whole team by default. Private is the exception, not the rule.
Written honestly, not aspirationally. If most of these describe you, the seat next to us will fit.
There's no single archetype at CreateLixir, but there's a shared disposition: care about the craft, care about each other, and enough curiosity to keep asking why. If those three are real, everything else is teachable.
People who read the papers, run the experiments, and rewire their opinions when the evidence says to.
People who'd rather ship a rough thing today and improve it tomorrow than argue about it for a week.
People who like ambiguity, and who chase the actual cause rather than symptom-swatting.
People who think of every surface as a decision — and who believe engineering and design are the same craft in different clothes.
People who care how the code reads six months from now and can hold a system in their head at the level that matters.
People who take the boring, careful path when the boring, careful path is the right one.
People who make the room around them stronger — through better questions, better feedback, and quieter egos.
People who feel physical discomfort shipping something they know is half-finished, and who won't be talked out of it.
A team of five behaves differently from a team of fifty, and both behave differently from a team of five hundred. Rituals that work today will need to evolve — and we plan to evolve them deliberately, not by accident.
What stays constant is the substrate: long-term thinking, ownership, craftsmanship, curiosity, simplicity, collaboration, respect, and continuous improvement. Those are the immovable parts. Everything else is scaffolding around them.
A culture of thoughtful engineering, curiosity, ownership, and long-term thinking. Practically: we ship carefully, we treat each other well, and we optimise for what the platform will look like on its fifth birthday.
Culture sits alongside the rest of the company section — same voice, same stance.
Builders, engineers, designers, and founders who care about craft and take the long view are always welcome to say hello — whether now or the next time you're between things.