Turning model output into reliable interfaces and systems, with a current focus on large-scale media processing pipelines and robust backends, processing hundreds of hours of content daily in production at Tessact.
Multi-tenant notifications platform
A notifications product that gives SaaS apps in-app inboxes and realtime updates out of the box. Teams send events through the backend SDK, drop in the React SDK on the frontend, and skip building the notification plumbing themselves.
Long-form to short-form, automated
A 12-stage media pipeline turns a two-hour podcast into 30 to 40 branded social clips in about an hour, with resumable orchestration, artifact-first stages, and parallel export tuned for production scale.
Video intelligence platform
Ask, search, and generate edits across large video libraries. Retrieval finds the right moments, ranking keeps narrative order intact, and model responses resolve into timelines editors can actually use.
Deep React years. Performance-critical UIs, chasing every dropped frame until the runtime stopped lying to me.
Pipelines, queues, infra. Started caring about queue depth more than bundle size.
Senior engineer at Tessact, working on large-scale video intelligence — processing hundreds of hours of content daily in production.