Senior AI Engineer – Search, Ranking & Applied ML

We are seeking a Senior AI Engineer – Search, Ranking & Applied ML, fluent in English, to offer their services as an expert based in Luxembourg.

Technical Requirements

• Strong expertise in Python
• Advanced experience with machine learning systems in production
• Strong experience with Learning-to-Rank (LambdaMART)
• Hands-on experience with: XGBoost, LightGBM (ranking objective preferred)
• Strong understanding of: feature engineering for ranking systems, structured + semantic feature blending, model evaluation (NDCG, MRR, Precision@K, Recall@K)
• Experience designing ML systems without historical click data

Search & Retrieval Requirements

• Experience building or working with: hybrid search systems (BM25 + embeddings), vector search architectures
• Practical experience with Vespa or a similar search infrastructure
• Strong understanding of: embedding generation pipelines, similarity search, and retrieval optimization
• Experience balancing recall vs precision trade-offs

LLM & Weak Supervision Requirements

• Experience using LLMs for: label generation, pairwise preference modeling, weak supervision workflows
• Experience with: LangGraph, LangSmith or LangFuse
• Practical knowledge of transformer fine-tuning
• Experience with the Hugging Face ecosystem
• Ability to design evaluation loops for LLM-generated supervision

Computer Vision & Document Understanding Requirements

• Experience building or improving OCR pipelines
• Familiarity with modern document understanding approaches: layout-aware models, transformer-based OCR systems
• Experience working with CV parsing and structured data extraction
• Ability to improve extraction quality beyond traditional Tesseract pipelines
• Understanding that parsing quality directly impacts ranking quality

System & Infrastructure Requirements

• Experience designing ML systems for: low latency, production reliability, cost efficiency
• Understanding of model serving architectures
• Experience building experimentation pipelines
• Ability to design evaluation frameworks from scratch
• Experience working with structured ranking signals: skills, seniority, location, experience alignment

Product & Design Mindset

• Strong product awareness - understands business impact of ranking quality
• Ability to design models aligned with recruiter workflows
• Thinks beyond model accuracy to: trust, interpretability, explainability
• Comfortable collaborating with frontend engineers to expose interpretable signals
• Makes pragmatic decisions instead of chasing unnecessary SOTA performance

Startup & Ownership Requirements

• Thrives in high-paced startup environments
• High autonomy, can design and ship systems independently
• Comfortable working with incomplete or evolving data
• Strong ownership of ranking quality end-to-end
• Able to define experiments and success metrics without predefined frameworks
• Balances research thinking with delivery speed
• Proactive, execution-focused, and highly accountable