**Cloudglue - Video Understanding Infrastructure**
Cloudglue is a **Y Combinator-backed startup** building developer APIs that turn video and audio into structured, searchable data. Think of us as the Stripe for video understanding - we handle the hard infrastructure (transcription, visual analysis, search, extraction) so developers can build on top of video without managing ML pipelines themselves.
Our team has shipped large-scale systems at Snapchat and Amazon, with work presented at AWS re:Invent, KubeCon, NeurIPS, ICCV, CVPR, and DEF CON. We process millions of minutes of video for customers building search, analytics, and automation products.
We’re a small, technical team where engineers have real ownership and direct impact on the product.
**The Role**
We’re looking for a **founding infrastructure engineer** to design and scale the backend systems that power Cloudglue’s video processing pipelines, search and retrieval infrastructure, and async job orchestration. You’ll be one of the first engineers on the team - this is a high-ownership role where you’ll shape the architecture and the engineering culture.
You’ll work on:
* Distributed video processing and async job orchestration
* Search and retrieval systems across video, audio, and text
* ML inference serving and model pipeline orchestration
* Storage, indexing, and compute infrastructure for large media collections
This is a **systems-heavy role** for someone who enjoys building reliable, high-throughput infrastructure and cares about getting the fundamentals right.
**What You’ll Do**
* **Distributed systems:** Design and operate async, fault-tolerant job execution systems that process thousands of hours of video reliably.
* **Search & retrieval:** Build and optimize search infrastructure across video, audio, and text - including vector search, re-ranking, and hierarchical retrieval.
* **ML infrastructure:** Own the serving and orchestration layer for ML models (vision, audio, language) in production.
* **Performance & reliability:** Profile and optimize throughput, latency, and cost across large-scale video workloads.
* **Production ownership:** Build systems that are observable, well-tested, and SOC2-compatible. You’ll own what you ship.
**What We’re Looking For**
**Required**
* 3+ years of backend or infrastructure engineering experience
* Track record designing and operating scalable distributed systems
* Strong proficiency in Python, Go, and/or TypeScript
* Ability to reason about performance, cost, and reliability tradeoffs
**Nice to Have**
* Experience with video/media processing or ML serving infrastructure
* Distributed systems or workflow orchestration (Temporal, Inngest, etc.)
* Vector search, retrieval systems, or ranking pipelines
* Cloud infrastructure (AWS/GCP), Docker/Kubernetes
* Experience with search or retrieval systems (vector databases like Milvus/Weaviate, ranking pipelines)
**Why Cloudglue?**
Video is the largest and most underutilized data source on the internet. Most software still can’t meaningfully work with it. We’re building the infrastructure to change that, and this role sits at the core of it.
If you want to work on:
* Hard distributed systems problems with real scale
* Search, retrieval, and ML infrastructure that doesn’t have off-the-shelf solutions
* A domain (video) where the infrastructure is still being invented
* A small team with massive leverage
…this is that role.
Visa sponsorship is not available for this role.
