Senior Site Reliability Engineer at Fanduel / Blip. Obsessed with building reliable and cost-effective systems and always looking for ways to improve things. In my free time, a foodie, an amateur photographer, and a homebrewing enthusiast.
Beyond the SELECT: Object-Centric Debugging with URNs and Safe AI Triage
In distributed architectures, the lion’s share of on-call triage time is spent opening SSH tunnels, hopping databases, and manually joining fragmented data. This post explores how URNs and the Model Context Protocol (MCP) enable a deterministic “Detective Agent”—complete with verifiable negative evidence logs, identity delegation, and object-centric safety constraints.
From O(n) to O(1): How We Made DataFrame Lookups 23x Faster
We optimize APIs. We scale databases. We architect systems for speed. But sometimes the biggest performance wins hide in plain sight — in a single function called thousands of times.
The Observability Pipeline: Taming Your Observability Costs
Stop paying to ingest useless data. An observability pipeline (using tools like OpenTelemetry Collector, Vector, or Redpanda Connect) acts as a smart router for your telemetry. We explore how this pattern lets you slash monitoring costs, enforce privacy by design, and break vendor lock-in.
Beyond the Key: The Power and Peril of End-User IDs
Unique API keys give you application-level identity. But for SaaS platforms, the real chaos comes from users. We dive into the pros and cons of using end-user IDs for abuse monitoring, cost attribution, and the significant privacy risks you must navigate.
The Hidden Dangers of Sharing API Keys Across Services
Reusing a single API key across multiple applications seems efficient, but it creates a nightmare for observability, rate limiting, and security. This post breaks down why this common anti-pattern is so dangerous and why unique keys are essential for robust systems.