Engineering Improving React page performance with AI agents How we used Cursor + Claude Opus 4.5 with a deterministic benchmark harness and a shared markdown file to cut p95 React frame time by ~76%.
Engineering How we cut engineering triage time by 80% with AI-powered observability At Metaview, we build AI agents for recruiting. Our stack runs on AWS Lambdas for most backend tasks, LangSmith for agentic workflows, Postgres on AWS, Datadog for logs and RUM,
Engineering From 200 Resumes to 5 Finalists in 24 Hours: Metaview's AI Screening Playbook Candidate screening is a signal problem, not a volume problem. The four moves that compress 200 inbound applications into 5 finalists in 24 hours, and the Metaview surfaces that make it repeatable.
Engineering Do Things That Don’t Scale - Metaview Wrapped In this post, we share the creation process of Metaview Wrapped, as an example of how Metaview experiments with ideas, and how we ‘Do Things That Don’t Scale’. [http://paulgraham.com/ds.html] At Metaview, we help companies run amazing interviews. We achieve
Engineering The Serverless Monolith How Metaview gets the structural simplicity of a monolith and the scaling independence of micro-services with a serverless architecture.
Engineering Syncing a Transcript with Audio in React How we built 60fps transcript-audio sync in our React web app in April 2022, using refs and direct DOM mutation to bypass React's render cycle.
