About the role
Nouvells is looking for Data Scientists who can transform complex datasets into meaningful insights, predictive models, and practical recommendations. You'll combine statistics, programming, machine learning, and business reasoning to solve data-driven problems.
As our Applied Data Scientist, you will own the algorithmic intelligence powering Nouvell’s core product features. You will not simply wrap external APIs; you will formulate mathematical and statistical solutions to complex platform problems building custom embedding spaces, fine-tuning small to medium open-weight language models, designing hybrid search/ranking engines, and developing adaptive user personalization systems. You will turn messy, multi modal interaction data into high performance models.
Key responsibilities
- Fine-tune and adapt open-weight language models (e.g., Llama, Mistral, Qwen) using parameter-efficient fine-tuning (LoRA, QLoRA), Supervised Fine-Tuning (SFT), and preference alignment (DPO, ORPO).
- Design robust intent-classification, entity-extraction, and structural output guardrail models integrated directly into application workflows.
- Mitigate retrieval bottlenecks in RAG pipelines by engineering context-compression algorithms, parent-child chunking heuristics, and semantic re-ranking strategies.
- Design and deploy adaptive recommendation systems and user affinity algorithms that model evolving multi-session user preferences.
Qualifications
- B.Sc in computer Science or any related field