Leveling the Playing Field: Building a 100% Offline Kubernetes Docs RAG Exam Assistant with Web AI, Gemma 4, and WebGPU
How to enable fair, zero-cost, air-gapped AI assistance in technical classrooms and...
First seen by Oriel on .
AI summary
The article describes Kubernetes WebAI RAG, an offline browser-based exam assistant that combines local documentation retrieval with on-device Gemma 4 generation to provide grounded answers without cloud APIs.
- Maintainers prebuild embeddings for 16,483 Kubernetes documentation chunks and a BM25 index, then distribute them with a sanitized local documentation mirror.
- At query time, EmbeddingGemma and BM25 search run in parallel, with Reciprocal Rank Fusion combining results before Gemma 4 E2B generates cited explanations and YAML.
- The student runtime uses WebGPU and requires a modern Chromium browser with hardware WebGPU enabled; the article targets Windows 11 x64 laptops and reports about 8 GiB of free disk space.
- The 2.01 GB model is cached in OPFS at a fixed localhost origin, while content security policies and citation handling restrict external access and tie citations to retrieved passages.
#retrieval-augmented generation #kubernetes #webgpu #gemma
Generated by gpt-5.6-sol. Check the original for details.
Metric trends
Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained
Reactions
Total public reactions reported by the source; distinct from points and likes.
- DEV Community: 10, latest observation ; +2 over observed period since
Comments
Comment count reported by the source.
- DEV Community: 0, latest observation ; 0 over observed period since
Aliases
| Source | External ID | URL |
|---|---|---|
| DEV Community | 4975841 | https://dev.to/gde/leveling-the-playing-field-building-a-100-offline-kubernetes-docs-rag-exam-assistant-with-web-ai-53bi Read discussion |