Title: The May 9th Blues: A Manager’s Perspective

---

May 2022 was a month of high expectations and relentless noise. As I sat down to reflect on my work, the tech landscape felt like a cacophony of buzzwords and deadlines, all vying for attention in my inbox. Here’s how it played out from my corner of the world.

The AI/LLM Tsunami

Since ChatGPT burst onto the scene, there’s been an insatiable demand for better AI infrastructure. We’ve been in overdrive, spinning up new instances and tweaking our models to handle more queries than ever before. It’s a bit like trying to drink from a firehose—there are just too many demands on our platform. The challenge is figuring out where to invest the limited resources we have.

Platform Engineering’s Mainstream Moment

Speaking of resources, platform engineering has truly come into its own this year. There was talk about FinOps and cloud cost pressures everywhere I turned. Our team has been grappling with how to balance performance with budget constraints, a task that’s both frustrating and endlessly rewarding. Debugging a slow-running query can feel like chasing ghosts in the database.

The CNCF Landscape

The CNCF landscape continues to overwhelm me. There are so many new projects and tools emerging all the time—it feels like trying to keep up with a moving train. We’ve been evaluating various options for our server-side applications, but the decision is never clear-cut. We’re torn between excitement and fear of making the wrong choice.

WebAssembly on the Server

WebAssembly (Wasm) has finally made its way onto the server side. It’s incredible to see how far it’s come, but also a bit daunting to implement in production. The learning curve is steep, and there are still plenty of edge cases we need to handle. For now, we’ve kept our Wasm experiments on the periphery, but I’m eager to dive deeper as more tools mature.

Developer Experience

Developer experience (DX) has become a real discipline. There’s been a lot of talk about how to make developers happier and more productive. At our company, we’re trying to streamline CI/CD pipelines and reduce friction in deployments. It’s not just about making the tools easier; it’s also about changing processes and expectations.

DORA Metrics

DORA (DevOps Research and Assessment) metrics are now normalized across teams. We’ve been using them as a way to measure our progress, but there’s always that nagging feeling of whether we’re doing enough. It’s hard to balance between pushing for more automation and maintaining the human touch in problem-solving.

A Personal Debug

A particularly memorable moment this month was debugging an issue with one of our microservices. We were seeing sporadic timeouts, which made tracking down the root cause a nightmare. After hours of tracing logs and profiling code, I finally realized it was due to a race condition in our Redis cache. Once fixed, the service became rock-solid again, but the journey was anything but straightforward.

The Great (and Not-So-Great) Debates

There were also some debates that got heated. For example, should we switch to server-side SQLite? On one hand, it’s a tried-and-true solution with great performance. On the other, it might not scale as well in our growing user base. We eventually decided to stick with what works for now, but the conversation was enlightening.

Wrapping Up

In conclusion, May 2022 was another month of intense activity and learning. The tech landscape is constantly evolving, and there’s always something new to figure out. As a platform engineer and engineering manager, it can be overwhelming at times, but that’s also what makes the job so exciting.

Until next time—keep coding!

---

Feel free to tweak or add any details you think are necessary!