Queue-it

Sam Dhar (@satyamdhar), Senior Staff Engineer and AI infrastructure leader at Galileo, talking with Jose Quaresma on the Smooth Scaling Podcast. Listen to Smooth Scaling wherever you get your Podcasts or in youtu.be/B07nzWlx_dI

you signed the data deal with your model provider. cool. the vendors helping you tune and evaluate that model are still sitting in the read path of everything it sees.

Sam Dhar (@satyamdhar), Senior Staff Engineer and AI infrastructure leader at Galileo, talking with Jose Quaresma on the Smooth Scaling Podcast. Listen to Smooth Scaling wherever you get your Podcasts or in youtu.be/B07nzWlx_dI

everyone's plan to cut ai costs is the same: self-host it. nobody can actually get the gpus. and the few who can still can't make the math work.

Chris Nesbitt-Smith on the Smooth Scaling Podcast with Jose Quaresma. on running kubernetes in production since v0.4, and why he'd reach for a managed service every time now. Listen to Smooth Scaling wherever you get your Podcasts or in youtu.be/9IVk8j8RJd4

Running your own Kubernetes cluster looks very clever right up until 4am. Every bit of operational responsibility you take on is a debt your successor pays.

Zaigham Sarfaraz, Engineering Manager at Queue-it, on the traffic pattern no autoscaler can touch — talking with Jose Quaresma on the Smooth Scaling Podcast. Listen to the Smooth Scaling Podcast wherever you get your Podcasts: youtu.be/CBHWUKFhoOM

"why don't they just use autoscaling?" it's the question every new queue-it hire asks. the answer changes how you think about traffic spikes.

Šimon Bučko, Senior Software Engineer at Queue-it, on why small teams should think twice before going horizontal — talking with Jose Quaresma on the Smooth Scaling Podcast. Listen to the Smooth Scaling Podcast wherever you get your Podcasts: youtu.be/CBHWUKFhoOM

Some "scalable microservices" can become distributed spaghetti. Vertical scaling is still a valid answer in certain scenarios.

10,000 log events all saying "error." But where? @irisdyrmishi from @MiroHQ on why tracing turns a two-hour investigation into a two-second one. New Smooth Scaling ep 🎧👇

When demand spikes, control matters. With @awscloud, we break down how Queue-it manages peak traffic so systems stay stable & access stays fair: ⚙️ Edge control with CloudFront 🤝 Fair access by design 🏗️ Architecture built for scale Check it out 👇 aws.amazon.com/blogs/apn/how-…

Every second of downtime during peak traffic is revenue lost. @irisdyrmishi from @MiroHQ on why observability is a scaling strategy — from cutting incident time to uncovering hidden cloud costs with eBPF. New Smooth Scaling ep 🎧👇

Check the full episode in Queue-it's youtube channel or your favorite podcast player

The "bots vs. humans" model is dead. 🤖👨 For years: bots = bad, humans = good. Simple. That's over. The new question isn't who is visiting your site — it's what they intend to do. Hans Skovgaard, CTO & CPO at Queue-it, on the shift every engineer needs to understand.

Hans Skovgaard, CTO & CPO at Queue-it, explains why the internet just changed — and what engineers need to do about it. 🎙️ Smooth Scaling Ep. 21 - Listen to the complete episode on your favorite Podcast Player.

Your autoscaling strategy won't save you anymore. AI agents don't browse. They don't hesitate. If something goes on sale at 10:00, every agent on the internet is there at 10:00:01 — and that burst will be more violent than anything we've seen before.

"Scalability is the ability to immediately sell faster than you did yesterday when you need to." — Art Huggard, after scaling Build-A-Bear's e-commerce from $6M to $70M. "The systems were on-prem, and I wouldn't recommend that to anyone."

Build-A-Bear's CEO: "How did you think of the waiting room idea?" Art Huggard: "I used to sell .22 ammunition the same way." Same demand spike problem. Completely different product. Best solutions are often borrowed from unexpected places.

When demand spikes to hundreds of orders per minute, small decisions have real operational impact 💥 In this clip, ecommerce consultant Art Huggard explains how a fast-selling product revealed scaling limits—and what the team adjusted to handle future launches more smoothly.