Built by people who ran on-call at scale
David Park spent four years as the on-call owner for a high-scale microservices platform in San Francisco. Most nights were quiet. Some nights were 3am pages for a CPU spike that resolved itself by 3:04am. Every month was a post-mortem that started with "why did we wake someone up for this."
In 2024, David and Priya Nair, a distributed systems engineer who had spent five years building internal alerting infrastructure at a travel technology company, decided to build the suppression layer they had both wanted for years. Marcus Okonkwo joined as Head of Platform Engineering, bringing experience managing alert runbooks across 200+ microservices at a data infrastructure company.
ObsrvHQ is bootstrapped. No investor pressure on roadmap direction, no enterprise sales motion to optimize for. If a feature is on this site, it works today. If you need something added, email us directly.
Three engineers. One problem they all lived with.
All three of us have been the person staring at a screen at 3am deciding whether a metric spike is real or noise. ObsrvHQ exists so that decision happens automatically, before the page fires.
Spent four years as on-call owner for a high-scale microservices platform in San Francisco. Built internal dashboards, wrote alert runbooks, and triaged too many 3am pages for self-healing CPU spikes. Started ObsrvHQ in 2024 to build the suppression layer that did not exist.
Five years building internal alerting infrastructure at a travel technology company: alert routing, threshold management, and eventually, a custom anomaly detection pipeline that never made it to production because it was too expensive to operate at scale. Builds the baseline engine at ObsrvHQ to make that work available without the overhead.
Former SRE lead at a data infrastructure platform, where he owned incident response and built alert runbooks for 200+ microservices. Knows exactly which alert categories produce actionable incidents and which are noise. Translates that experience into how ObsrvHQ's default rule schema is structured.
How we build
The product page describes what ObsrvHQ does today. If a capability is listed, you can use it right now. Features we have not built yet are not on the marketing site. Roadmap promises are a different kind of noise.
Alert fatigue is not a productivity problem. It is a human problem. Engineers who lose sleep to rotation noise, who start ignoring pages because the signal-to-noise ratio has trained them to, are not a talent issue. They are a tooling issue that the tooling industry has been slow to acknowledge.
Not the manager who approves the tool purchase. Not the finance team reviewing SaaS spend. The on-call engineer at 3am, reading a metric value and deciding whether this is worth waking anyone else up for. Every product decision runs through that lens.