Clips AI — Video-to-Shorts Pipeline
3 min readLong streams need automated transcription, LLM clip discovery, and vertical renders without blocking HTTP or duplicating work across retries.
Node.jsNestJSNext.jsPostgreSQL
// systems engineer
Sergey Orsik
AI / Systems Engineer
I build AI systems, realtime pipelines, and distributed infrastructure — proof-of-work over pitch decks.
orsik@portfolio ~
Type a command or click a chip below. Try: whoami
// about
Engineering focus
Not a life story — how I think about systems.
- →Systems thinking: design for failure, observability, and clear boundaries.
- →AI engineering: memory tiers, agent orchestration, inference routing.
- →Realtime: WebSockets, event streams, sub-100ms delivery paths.
- →Distributed: queues, caches, and data planes that scale horizontally.
// projects
Selected work
Problem → architecture → stack. Proof of engineering ability.
Video-first Dating App
5 min readUser-generated profile media must be validated, moderated, and transformed without blocking HTTP or overloading the API tier.
Spring Boot (JVM API)Go (media worker)Flutter (mobile + web)PostgreSQL
AI Agent Orchestration Platform
1 min readTeams needed multi-step agents with tool use, retries, and audit trails — without a spaghetti of cron jobs.
PythonTemporalRedisPostgres
// engineering notes
How I think
Short writeups on architecture and system design.
2 min read
AI Run Audit Trail and Prompt Versioning
Every LLM call creates an ai_runs row with input hash, prompt version FK, token usage, and validation errors for replay and debugging.
5 min read
Event Contract — Media Optimization (API ↔ Worker)
Version-agnostic Kafka message schemas, topic naming patterns, delivery semantics, and consumer mapping rules.
// contact
Let's build
Founders, recruiters, startups — reach out.