Research Engineer, Multilingual Semantic Search

TGScope is hiring a Research Engineer to evaluate and improve its multilingual semantic search. The role focuses on benchmarking relevance, measuring retrieval quality, improving search methods, testing models within latency constraints, and documenting results.

Responsibilities:

• Build a relevance benchmark for channel search in 6+ languages: query sets, relevance judgments, nDCG and recall@k
• Measure retrieval quality by language and query type and analyse failure cases
• Improve embedding inputs, vector weighting and the fusion of semantic and lexical results
• Test stronger embedding models and re-rankers within a CPU-only latency budget
• Document the method and results

Requirements:

• Experience in information retrieval, search or NLP engineering
• Hands-on work with dense retrieval and hybrid search in production
• Strong Python
• Experience with a vector database (Qdrant, Milvus, Weaviate or similar)
• Experience with multilingual embedding models
• Written English
• Model export and optimisation with ONNX
• Experience with low-resource languages
• Published IR or NLP work, open-source contributions