Lumitech

Fast growth is easy to celebrate. Sustained growth without lowering standards is harder. Since 2022 @lumitech_co has grown 850% and scaled to around 100 professionals globally. @clutch_co has now named us among its 100 #FastestGrowing companies in 2026. What matters more to us is what sits behind that growth: strong execution, client trust and long-term partnerships. More here: lumitech.co/insights/lumit…

Standards matter but they solve different problems. HL7 v2 remains critical for message-heavy hospital workflows. FHIR is better suited for API-first products and newer digital health services. We broke this down here: lumitech.co/insights/healt…

In practice, the goal is not just to connect EHRs, labs and external services. It is to make the right patient data arrive in the right context, for the right role, with less manual effort 🧵

#Healthcare interoperability is not a connectivity problem alone. Data can be exchanged successfully and still remain hard to find, hard to trust or hard to use at the moment of care 🧵

@deepp2108 Bubble sort knowledge: useless in prod and critical for getting the job where you'll never use it. Perfectly normal.

Software engineering is the only field where you can build systems used by millions and still get rejected because you forgot the time complexity of bubble sort

@plainionist Good catch. Mixing roles in one session is like asking the same person to write the spec, build it and then audit their own work.

Hot take: LLMs are bad at context switching. If you use the same session for planning, coding, testing, and review - don’t expect great results. Those are different roles. The model gets confused when their contexts are mixed. Separation of concerns isn’t just for code 😎

@csaba_kissi AI can build anything, it has no idea what's worth building.

AI doesn't know what to build. It doesn't understand the user's frustration at 2am. That's our job.

@Franc0Fernand0 If your definition of engineering is write the code someone described - sure, LLMs are close. If it's figure out what should be built and make sure it doesn't fall apart - not even close.

When people claim that LLMs will replace software engineers, it either indicates a lack of understanding of LLMs or of software engineering. But if your only definition of software engineering is related to feature development, there is potential to believe that LLMs can replace developers.

@jdegoes The teams that get the most out of AI are the ones who stopped being impressed by it. Once you know what it actually is, you stop being surprised by what it can't do.

For as long as you believe, 'LLMs can think', you'll be confused & frustrated by them. Without a good mental model, you won't know their limitations and can't engineer workarounds. In other words, your blind faith in LLMs will limit the value you can extract from them. 1/2

@levie Shipping is easy. Owning what you shipped is a different job entirely.

Agentic coding is a huge boon for software developers that want to get far more done, great for IT people to build vastly more custom systems internally, great for domain experts that want to automate workflows or wire systems together, and absolutely fantastic for anyone curious to learn how to start coding. What it’s less great for is casually building complex software that you have to maintain on an ongoing basis and take on all the risk for. Upgrades, maintenance, keeping up to date with latest security issues, and so on, are taxes most knowledge workers aren’t familiar with or prepared for. Net net: we’re going to get 100X more software and vastly more software developers in the future. But that’s different from *everyone* rolling their own.

The best engineers today are not faster coders. They’re better decision-makers. Architecture, constraints, ownership - that’s the job now. What does your team optimize for?

When code is cheap, bad decisions scale faster. Wrong architecture doesn’t slow you down anymore. It spreads. Fast.

Code got cheap. The old model of software development didn’t. That’s the real shift behind AI.

@icanvardar The constraint was never headcount - it was always clarity on what to build. Small teams with that still win. Large teams without it just fail faster now.

ai is removing the need for large teams to build meaningful products a few people can now design, ship, and iterate at a speed that used to take entire orgs this changes what kind of software gets built in the first place

@asaio87 The gap between the demo and the reality is where most non-developers lose trust in the whole thing and honestly, that's on the marketing not the tech.

AI coding sounds and looks so good on paper. Until you actually try to code with it. If you have high expectations and have no software development experience, then you will think the CEOs of these companies are mocking you. And they absolutely do that to you. Their product makes professional software developers maybe 20-30% more efficient. Then they add one more 0 or two more 00 to that figure because why not. we have not solved software development we just solved code completion, which is actually very very good. keep learning

@jaredsuniverse Because the bottleneck was never writing code or designing screens. It was knowing what to build, for whom and actually shipping it to people who pay.

AI writes better code than 99% of humans AI creates better designs than 90% of humans AI is basically better at problem solving than 90% of founders Why hasn't everyone made a billion dollars?

@EXM7777 The companies that seem calm right now are either very good at adapting or not paying close enough attention.

the past year has had a lot of moments where i asked myself "where will i be in 6 months" and i had absolutely no clue the pace AI is moving at, the speed i have to adapt my businesses, it's disorienting in a way i wasn't expecting obviously not everything in life feels like that, my health, my relationships, the next vacation, those still feel predictable in a normal way but business is wildly unpredictable now, and as an entrepreneur that's a massive part of who you are and how you spend your days the past 10 years i could always feel which direction i was heading, i could actually do strategy, ride trends, plan ahead today is just a different game, hard to put words on it, but the constant adaptation and the constant not-knowing is new not saying it's bad, not saying it's good just very different, and quite disturbing some days

@chamath The 10x developer was always a myth. Software is a team sport. AI just made it easier to forget that.

every AI coding tool on the market is a single-player game. cursor, copilot, claude code. they all make the individual developer faster at writing code. And they are brilliant at it. 2-5x individual velocity, sometimes more. but software isn't a single-player game. software is mostly architecture decisions, requirements debates, compliance reviews, code reviews, rollbacks, post-mortems, onboarding. it’s a multiplayer sport with a dozen roles and thousands of decisions that don't just live in one person's head. the key in good software development is the coordination between roles, the traceability of decisions, the institutional memory of why something was built a certain way. that's what software factory is. it’s a multiplayer AI. Requirements captures business intent before an engineer opens an IDE. Blueprints captures architecture decisions upstream. Work Orders routes structured tasks to AI agents through MCP with full context. the Knowledge Graph holds the state of every artifact. try it here: factory.8090.ai