Multilingual RAG for a European Customer-Support Knowledge Base
Overview
What this challenge is about.
You receive 6,000 documents in 4 languages (mix of FAQs, parts catalogs, repair procedures) plus 120 labeled queries (30 per language) with gold source documents. Build a multilingual RAG: use a multilingual embedding model (e.g., BAAI/bge-m3 or multilingual-e5-large), retrieve cross-lingually, optionally rerank with a multilingual cross-encoder, generate in the query language with citations to original-language sources. Evaluate per-language Hits@5, faithfulness, and answer-language fidelity (the answer is fully in the query language). Success is per-language Hits@5 above 80 percent and answer-language fidelity above 98 percent for all 4 languages.
The Brief
What you'll do, and what you'll demonstrate.
Build a multilingual RAG that retrieves across 4 languages and answers in the query language with original-language citations, meeting per-language quality bars.
Earning criteria — what you'll demonstrate
- Apply multilingual embedding models to cross-lingual retrieval
- Design per-language evaluation that surfaces minority-language weaknesses
- Manage citation hand-off between source language and answer language
- Communicate readiness to an enterprise customer's leadership
Program Fit
Where this fits in your program.
Sharpens the same skills your degree expects you to demonstrate.
Skills
Skills you'll demonstrate.
Each one shows up on your verified credential.
Careers
Roles this prepares you for.
Real titles. Real skill bridges. Pick the one closest to your trajectory.
NLP Engineer
Multilingual RAG with cross-lingual retrieval and per-language evaluation is core NLP-engineer work in any global enterprise AI deployment.
This challenge sharpens
- multilingual-rag
- cross-lingual-retrieval
- multilingual-embeddings
AI Solutions Architect
Translating enterprise multilingual requirements into a pipeline architecture plus a launch-readiness memo is exactly the AI-solutions-architect skill set.
This challenge sharpens
- multilingual-rag
- citation-handling
- rag-evaluation
AI Engineer
Wiring multilingual embeddings + reranker + multilingual generation into a customer-deployable pipeline is core AI-engineer work.
This challenge sharpens
- multilingual-rag
- python
- citation-handling