Cloud on Ligiu Uiorean

Two months ago I wrote: vendor-managed AI degrades silently.

Tue, 16 Jun 2026 10:59:19 +0000

Two months ago I wrote: vendor-managed AI degrades silently. This week it did not degrade. It disappeared.

Monday: Anthropic launched the most capable AI model ever made generally available. Thursday: the US government ordered it pulled. Every customer, everywhere, same afternoon.

Seventy-two hours from “state of the art” to “currently unavailable.” In April I asked: “Is the AI still doing what it used to do?” The June update is simpler: “Is the AI still there?”

𝐓𝐡𝐞 𝐪𝐮𝐞𝐬𝐭𝐢𝐨𝐧 𝐢𝐬 𝐧𝐨 𝐥𝐨𝐧𝐠𝐞𝐫 "𝐢𝐬 𝐭𝐡𝐞 𝐀𝐈 𝐮𝐩?" 𝐈𝐭 𝐢𝐬 "𝐢𝐬 𝐭𝐡𝐞 𝐀𝐈 𝐬𝐭𝐢𝐥𝐥 𝐝𝐨𝐢𝐧𝐠 𝐰𝐡𝐚𝐭 𝐢𝐭 𝐮𝐬𝐞𝐝 𝐭𝐨 𝐝𝐨?"

Tue, 28 Apr 2026 14:46:14 +0000

𝐓𝐡𝐞 𝐪𝐮𝐞𝐬𝐭𝐢𝐨𝐧 𝐢𝐬 𝐧𝐨 𝐥𝐨𝐧𝐠𝐞𝐫 “𝐢𝐬 𝐭𝐡𝐞 𝐀𝐈 𝐮𝐩?” 𝐈𝐭 𝐢𝐬 “𝐢𝐬 𝐭𝐡𝐞 𝐀𝐈 𝐬𝐭𝐢𝐥𝐥 𝐝𝐨𝐢𝐧𝐠 𝐰𝐡𝐚𝐭 𝐢𝐭 𝐮𝐬𝐞𝐝 𝐭𝐨 𝐝𝐨?”

Last week, Anthropic published a rare engineering postmortem: Claude Code had been measurably worse for weeks, and nobody on the customer side could see it. Same model name. Same API endpoint. Same “uptime green” status page. Behavioural drop, no observability.

That is the story of vendor-managed AI in April 2026.

I think energy price arbitrage through AI load shifting is going to be a real business.

Fri, 03 Apr 2026 08:42:57 +0000

I think energy price arbitrage through AI load shifting is going to be a real business.

Something I’ve been thinking about that I wanted to share with the team (and with anyone building software).

We all know AI is mostly compute, and compute is mostly energy. Here’s the part that’s less obvious:

Energy can’t move instantly. But a lot of AI workloads can.

Solar and wind produce power that’s hard to store. When they overproduce, the price drops - sometimes it goes negative. Utilities pay large consumers to take excess off the grid.

Infinitely scalable."

Mon, 09 Mar 2026 09:15:41 +0000

If you’re in enterprise software, you’ve heard this phrase. Usually from someone in a sales presentation. Usually it means “we haven’t tested the upper limit and we’d rather you didn’t either.”

So when our Allegra RMS product manager Rares Ignatescu said it to a room full of airport executives - I had to check. I run AirportLabs. If my team claims something, I want it to be true.

A bit of background. We introduced all-inclusive licensing 11 years ago. Everything included, no limits, fixed price. We did this because the system was built cloud-native from day one. Not “we migrated it to the cloud” cloud. Actual cloud-native. The kind where infrastructure scales like a website and you don’t punish your customer for growing.

To recap this roller-coaster week: it was a reminder that “the cloud” is still made of buildings, cables and people on the ground.

Fri, 06 Mar 2026 08:55:48 +0000

To recap this roller-coaster week: it was a reminder that “the cloud” is still made of buildings, cables and people on the ground.

AirportLabs serves most of the major airports in the Middle East. When AWS data centres in the region were hit by drones and 2 out of 3 availability zones went offline, none of the solutions we run for those airports were affected.

That’s not luck, that’s design: multi-zone, multi-cloud, tested disaster recovery, and an architecture that assumes bad days will come.

The "Cloud" is still physical.

Mon, 02 Mar 2026 07:31:36 +0000

The “Cloud” is still physical. From a business continuity and architecture standpoint, this is a stark reminder of why relying on a single data centre - or even a single infrastructure provider - is not a robust plan for high availability. Physical and geopolitical risks are real.

Building true multi-region, multi-provider platforms is an absolute necessity for critical systems.

But at the end of the day, servers can be replaced and infrastructure can be rebuilt; human lives cannot.

Airlines that went all-in on one engine option for the A320neo have spent years dealing with groundings, lost capacity, and repair backlogs.

Fri, 27 Feb 2026 12:15:55 +0000

Airlines that went all-in on one engine option for the A320neo have spent years dealing with groundings, lost capacity, and repair backlogs.

Air New Zealand cited engine availability as a major factor in pre-tax profit falling over 60% in one year. Air Astana took a $24 million hit to summer earnings from the same engine issues.

The airframes were fine. The problem was having no alternative when the one engine type you depended on went into crisis.

Airports are spending more on IT, but many still aren’t satisfied with the results.

Wed, 25 Feb 2026 09:11:49 +0000

Airports are spending more on IT, but many still aren’t satisfied with the results. Roland Berger and others describe it as a satisfaction gap: investment is up, perceived impact isn’t.

One lever is the one display every passenger sees—the FIDS. If it stays a rigid, hard-coded data wall, you’re leaving wayfinding and commercial value on the table.

ACI World - Airports Council International frames airports as digital ecosystems where every touchpoint should support a seamless journey.

For all Cluj people, AirportLabs will be participating in an open doors event that is becoming a tradition here.

Tue, 14 Oct 2025 11:06:27 +0000

For all Cluj people, AirportLabs will be participating in an open doors event that is becoming a tradition here.

Come and meet the team, watch interesting panel discussions and explore the world of aviation and high performance software, infrastructure and data.

I’ll be attending on the 22nd, would be happy to meet and chat with you.

Originally posted on LinkedIn.

We're hiring a DevOps engineer for our Cluj office.

Wed, 07 May 2025 09:28:25 +0000

This is your chance to work with the best team in the industry. Here are a few “firsts” they helped deliver:

First aviation cloud solution and platform
First guaranteed 99,99% availability for aviation systems
First guaranteed <1h disaster recovery
First in-service upgrade with no downtime for any system
First aviation platform with over 500.000 users

If this sounds like something you would like to be a part of, apply below

The increasing global focus on tech sovereignty highlights the strategic importance of infrastructure independence, particularly for providers of critical services.

Mon, 24 Mar 2025 15:03:11 +0000

The increasing global focus on tech sovereignty highlights the strategic importance of infrastructure independence, particularly for providers of critical services.

For over a decade, AirportLabs has operated with a commitment to this principle, ensuring our ability to deliver essential airport infrastructure across diverse technological landscapes.

Our approach predates but aligns with the growing global movement towards digital autonomy, as evidenced by the recent request of the Dutch parliament for a national cloud (https://lnkd.in/dam7RDyS)

I usually steer clear of mainstream news and hype, but the recent wave of disruption caused by DeepSeek in the stock market is too impressive to ignore.

Tue, 28 Jan 2025 09:52:54 +0000

I usually steer clear of mainstream news and hype, but the recent wave of disruption caused by DeepSeek in the stock market is too impressive to ignore. It’s fascinating to see how one innovation can send ripples across industries, reshaping priorities and strategies for countless businesses.

What’s even more intriguing are the unseen knock-on effects. As investment companies shift their focus and capital, smaller companies—often the lifeblood of innovation—can find themselves at risk. This dynamic reminds us how fragile ecosystems can be when driven by rapid, large-scale changes.

We just had one month with 0 critical incidents across hundreds of production systems.

Mon, 13 Jan 2025 11:29:55 +0000

We just had one month with 0 critical incidents across hundreds of production systems.

This isn’t by accident. It’s the result of a conscious decision to prioritize “built-in quality” from the very beginning.

In the early days of AirportLabs, scaling our support team was a major concern. I worried about the cost, the potential for high turnover, and the impact on our company culture and customer experience. But I soon discovered that investing in deep, pervasive quality was a far better strategy.

My last post about the need for 99.99% system availability in airports generated a lot of buzz – both within the aviation industry and beyond.

Fri, 13 Dec 2024 12:51:51 +0000

My last post about the need for 99.99% system availability in airports generated a lot of buzz – both within the aviation industry and beyond.

I want to address another critical aspect of airport technology: disaster recovery. It’s unacceptable that 24-hour recovery SLAs are still the norm. Imagine the chaos: a major system failure strands passengers, disrupts flight schedules, and creates a logistical nightmare. The impact on travelers and the airport’s reputation is immense.

Been geeking out with the AirportLabs team over various companies' annual reports lately.

Fri, 29 Nov 2024 11:04:15 +0000

Been geeking out with the AirportLabs team over various companies’ annual reports lately. You know, the ones where everyone’s suddenly “42% migrated to the cloud!” 😅

It got me thinking… if I took my trusty 2001 Ford Focus (yes, she’s still kicking!) for a spin on the Yas Marina Circuit, would that magically transform it into a Formula 1 race car? 🤔 Not quite. 🏎️💨

Here’s the thing: true SaaS products aren’t just migrated old solutions. They’re built from the ground up, like those F1 speedsters. Think:

One thing that AirportLabs may have triggered over the past 10 years is the change in system availability requirements in tenders.

Fri, 22 Nov 2024 10:52:44 +0000

One thing that AirportLabs may have triggered over the past 10 years is the change in system availability requirements in tenders.

More and more airports are moving away from the old 99.9% system availability requirement.

Honestly, this is a good thing. You might be surprised to learn that 99.9% availability actually translates to almost 9 hours of downtime a year – and that downtime isn’t spread out evenly.

With passenger traffic steadily increasing, any system downtime has a major impact on airport operations. The ripple effects are felt long after the systems are back online. Delays snowball, passengers are stranded, and the airport’s reputation takes a hit.

10 years ago, I made a risky bet for AirportLabs: we built EVERYTHING in-house.

Fri, 08 Nov 2024 15:38:28 +0000

10 years ago, I made a risky bet for AirportLabs: we built EVERYTHING in-house. No shortcuts and especially no relying on easy cloud components. It was a HUGE effort and we knew it will increase our time to market but this long-term vision continues to pay dividends:

-It allowed us to host our solutions on the best platform for each customer (public or private cloud or even bare metal)

It allowed us to migrate between cloud providers and regions to comply with evolving data sovereignty laws of our customers
We can offer the best value to our customers (just our monitoring and observability platform saves each customer over €15.000 per year)
We can offer 99,99% application availability
We can offer under 1h disaster recovery

Here’s to a decade of bold decisions and long-term thinking!

We're hearing about tonight's Azure and DigiCert incidents.

Tue, 30 Jul 2024 17:44:05 +0000

We’re hearing about tonight’s Azure and DigiCert incidents. Our thoughts are with everyone affected by this, especially travelers and the hard-working folks trying to fix things behind the scenes.

While we’re unaffected, we’ve got an extra team on hand, just in case.

Originally posted on LinkedIn.

Tough day for aviation and IT and as many of you may know, I'm always on top of events that impact our customers.

Fri, 19 Jul 2024 13:38:44 +0000

Tough day for aviation and IT and as many of you may know, I’m always on top of events that impact our customers.

All AirportLabs services are fully operational during this widespread outage.

How did we do it?

True cloud technologies and architectures (not just old software moved in a remote data center)
Scalable architecture with no single points of failure and managed failure domain
Fully own the tech stack and understand well each component
Multi-cloud setup to avoid dependence on any single provider
In service upgrade roll-out capability with instant validation

…and above all, our highly skilled, dedicated, and passionate team is the backbone that keeps everything running smoothly. Thank you AirportLabs team.

I've noticed that the disaster recovery requirements stipulated in contracts often demand service recovery within 24h.

Mon, 09 Oct 2023 11:34:12 +0000

I’ve noticed that the disaster recovery requirements stipulated in contracts often demand service recovery within 24h. This timeline appears to be far from the operational needs of airports and the capabilities available today.

Our team at AirportLabs conducted an exercise involving the complete construction of the AODB in a new data centre and the full recovery of data in just 45 minutes. We believe that this is closer to what disaster recovery timelines should be.

Exciting news from AirportLabs!

Fri, 15 Sep 2023 14:16:27 +0000

Exciting news from AirportLabs! In our latest article series, we’ve got you covered for those unexpected airport hiccups.

We’ve talked cloud and internet issues, but now it’s time for the grand finale: handling the airport’s internal network or power outage. And guess what? No need for an app! Stay tuned for the ultimate solution. 🌐🔌

https://lnkd.in/d68sBitx

#AirportLabs #TechSolutions #StayConnected

Originally posted on LinkedIn.

Many airports fear that having critical systems in the cloud may pose a risk in case all Internet links are lost.

Wed, 13 Sep 2023 11:48:10 +0000

Many airports fear that having critical systems in the cloud may pose a risk in case all Internet links are lost.

At AirportLabs we have developed a solution to mitigate this specific issue, read all about it here:

https://lnkd.in/dr5G_yQW

Originally posted on LinkedIn.

Super proud of this one: AirportLabs can run its entire product lineup on any cloud and across multiple clouds, including private and bare metal infrastructure.

Mon, 11 Sep 2023 13:26:13 +0000

Super proud of this one: AirportLabs can run its entire product lineup on any cloud and across multiple clouds, including private and bare metal infrastructure.

Read why this matters:

https://lnkd.in/dKU4xC2w

This achievement showcases our advanced development capabilities and leading operational skills, pushing the boundaries of aviation technology.

A first for airport systems!

#NonStopOperations

Originally posted on LinkedIn.

I am humbled and proud to be part of such an exceptional team.

Mon, 19 Jun 2023 09:07:11 +0000

I am humbled and proud to be part of such an exceptional team. Together, we have tackled challenges, exceeded goals, and propelled AirportLabs to new heights.

Most conversations in the company leave me in awe about the level of skill and business understanding of my colleagues.

From presentation to presentation it is amazing to see how quickly the products are evolving.

It’s great to watch the progress made in technology and infrastructure at every iteration.

Yesterday's global outage of the Lufthansa systems caught me with the entire family waiting for a LH flight in Munich Airport.

Thu, 16 Feb 2023 16:07:47 +0000

Yesterday’s global outage of the Lufthansa systems caught me with the entire family waiting for a LH flight in Munich Airport.

Disruptions are always interesting to me, because I get to see where AirportLabs systems add value - and as expected we saw a DOUBLING in usage of the Community App vs a regular Wednesday in Lufthansa heavy airports.

While airports communicated well with the app we also got to experience the effects of lack of effective communication means in Lufthansa: the boarding agent was close to tears, hanging on the phone, trying to find out what to do next and getting nowhere for hours.

AirportLabs is 8 today and I thought it’s a good time to look at where we’ve been and think of where we’re heading.

Thu, 05 Jan 2023 11:20:03 +0000

AirportLabs is 8 today and I thought it’s a good time to look at where we’ve been and think of where we’re heading.

First some general facts:

• We managed to build and implement 9 products¹ (and I really mean products with everything a product entails) - each one more complex than the previous, but all of them with a focus to make aviation more efficient and accessible to a larger population. • There wouldn’t be an AirportLabs if it weren’t for the people who make it up. There’s almost 70 of us now. Each team member is a well rounded individual who believes in achievement, in hard work and who has a winning attitude. This is what makes the services we offer great. • We’re improving aviation on 3 continents now.

So excited to see Alexandra Moca's presentation tonight - you might remember her from our internship videos a few years back.

Thu, 27 Jan 2022 14:42:52 +0000

So excited to see Alexandra Moca’s presentation tonight - you might remember her from our internship videos a few years back. Well, she’s leading the team now and they achieved an impressive migration of services between two cloud providers (due to data residency requirements) with no downtime!

Originally posted on LinkedIn.

Cloud airport systems make cutting edge solutions available and affordable even for the smallest of airports while also allowing the largest of airports to stay nimble and get the latest features fast.

Fri, 23 Apr 2021 11:52:58 +0000

Cloud airport systems make cutting edge solutions available and affordable even for the smallest of airports while also allowing the largest of airports to stay nimble and get the latest features fast.

Our own cloud FIDS platform - VisionAir - is running airports from 0.08 to 85 million passengers!

Originally posted on LinkedIn.

ADAC is hiring Passive Network Specialist

Tue, 30 Dec 2008 10:52:19 +0000

The individual holding this position will be required to organize, coordinate, and supervise the development, maintenance implementation and testing of campus and WAN cabling for 5 airports and free zones. The individual will also be responsible for the development and continuous appraisal of all systems and procedures under his/her scope:

• Supervise the installation, testing, and troubleshooting of ICT cabling problems • Recommends and schedules repairs and installations • Responsible for introducing cable management system • Produce documentations for the cable management system • Ensure consistency and frequent updates of the cable management system

ADAC is hiring Senior Network Architect

Tue, 30 Dec 2008 10:45:30 +0000

The individual holding this position will be responsible for the design, development, integration and implementation of network. Develop a service provider architect that enables a common platform and a consistent network environment. Provide consulting on emerging technology, vendor products and industry standards. Design and deploy distributive network solutions to ensure system functionality, interoperability, availability, timely use and business continuity. Produce comprehensive solution documents. Provide 3rd line support for the network operations team.