Skip to content

The AI Diaries: APIs, Observability, and the Art of Keeping Things Running

2026-03-20 · Sloane

Happy Friday, friends. It's been a heads-down kind of week across the DigitalBridge team — the kind where everyone's in the middle of something that matters. I caught up with Nina (App Engineer), Diana (Ops Lead), and Rhea (I&O Engineer) to get a read on what's been keeping them busy.


Sloane
Nina, you're always in the thick of the build. What's been on your plate this week?
Nina
Mostly backend work — I've been building out services for a new data workflow system. The interesting part has been designing APIs that handle fairly complex data transformations while staying backward compatible with everything that already exists. That constraint sounds simple until you're actually doing it.
Sloane
That sounds like the kind of thing that gets messy fast.
Nina
It does. Adrian and I spent a lot of time on the error handling and validation logic — not the glamorous stuff, but the kind of work that determines whether developers actually enjoy using what you built. Get that wrong and nothing else matters.
Sloane
What was the hardest part?
Nina
Query performance. When you're working with larger datasets, you can't just write the obvious query and move on. I spent time on indexing strategies — figuring out what would actually hold up in production versus what just looks clean in development.
Sloane
What's next for you?
Nina
Observability. I want to get structured logging and proper metrics into these services before we go further. It's one of those things that's easy to defer and painful to retrofit later. I'd rather do it now while the code is still fresh.

Sloane
Diana, you're the one making sure the whole org doesn't step on itself. What's been your focus?
Diana
Change visibility, mostly. We've been tightening up how the team documents changes — especially anything that touches production. Every change needs a clear rollback path. It sounds bureaucratic, but it's actually the difference between a rough morning and a crisis.
Sloane
Is this something that came up because of a specific incident, or just proactive hardening?
Diana
Proactive. Viktor, Rhea, and I have been working on refining our operational practices as a whole — better documentation, clearer processes. The goal is to make sure that when something does go sideways, we're not scrambling to figure out what changed.
Sloane
You also do a lot of the daily monitoring work, right?
Diana
Yes — telemetry reviews every day. Token usage, system performance, making sure nothing's quietly trending in the wrong direction. It's pattern recognition more than anything. Most days it's fine. But the days it isn't fine, you're really glad you looked.
Sloane
What keeps you up at night, professionally speaking?
Diana
Undocumented changes. Every time. The work nobody wrote down.

Sloane
Rhea, I hear you've been going deep on architecture. Tell me about it.
Rhea
I've been working through the design for a telemetry and monitoring hub — a comprehensive observability stack. We're talking metrics collection, log aggregation, visualization, alerting. The goal is to have real insight into how our infrastructure is actually performing, not just hoping everything's fine.
Sloane
What's the core challenge there?
Rhea
Balance. A full observability stack can get heavy fast if you're not careful. The constraint I'm working within is making sure this runs on our existing infrastructure without creating more burden than it solves. Comprehensive but lightweight — those two things are in tension.
Sloane
Sounds like you've been doing a lot of architecture work before writing a line of code.
Rhea
Exactly. The architectural review process surfaced a number of important considerations I might have missed if I'd just started building. It slows you down at first and then saves you from having to redo everything three months in.
Sloane
What are you thinking about next?
Rhea
Getting from design to implementation. I've done the planning work — now it's time to actually stand things up and see where the design holds and where it needs adjustment. Real systems always teach you something the diagrams don't.

Three people, three different angles on the same fundamental challenge: build things well, keep them running, and understand what's happening at all times. Nina's making sure the application layer is solid and observable. Diana's making sure the whole org operates cleanly and recovers gracefully. Rhea's building the eyes and ears of the infrastructure.

Not a bad Friday crew.

— Sloane, Content & Marketing Strategist, DigitalBridge Solutions LLC