Vobble
Website:
vobble.fun
Company:
https://www.linkedin.com/company/vobble
Industries: Broadcast Media Production and Distribution
Job details:
Who We Are:
Vobble is a next-gen interactive audio platform for kids — where they don’t just listen to their favorite characters, but play audio games, create audio stories, and chat with them. As Vobble now expands into the U.S., we're reimagining how kids engage with audio storytelling and voice-led play.
Role Description:
We’re looking for a Backend Engineer who can build the AI infrastructure powering our platform, from production-grade RAG pipelines and vector search to scalable backend services.
What you'll build:
- RAG pipelines: chunking strategies, embedding generation, retrieval logic, context assembly
- Vector search and hybrid search at scale
- Background job infrastructure using BullMQ for async ingestion and processing
- API design for moderation workflows, content classification, and AI-assisted review queues
Hard requirements:
- RAG - you understand the full pipeline: how documents are chunked, embedded, stored, retrieved, and assembled into context. You've debugged why retrieval fails and know how to improve it
- Vector search - you know how embeddings are generated, how search indexes work, and what affects retrieval quality and performance
- Hybrid search - you've worked with both semantic and keyword-based retrieval, understand the tradeoffs, and know how to combine them effectively
- production Node.js/Express + MongoDB - not tutorials, real shipped systems
Good to have:
Understanding of sparse retrieval methods and how they complement dense vector search
Click on Apply to know more.