- Company Name
- Proximity Works
- Job Title
- Senior Data Scientist - LLMs, RAG & Multimodal AI (Remote | Immediate joiner)
- Job Description
-
**Job Title**
Senior Data Scientist – LLMs, RAG & Multimodal AI
**Role Summary**
Lead the design, fine‑tuning, and production of large language models and multimodal systems that combine language, vision, and retrieval capabilities. Build and optimize retrieval‑augmented generation (RAG) pipelines, manage model distillation, and ensure high‑quality, low‑latency inference at scale.
**Expectations**
- Deliver production‑ready LLM + RAG systems for global search and discovery applications.
- Demonstrably reduce hallucinations and improve grounding accuracy.
- Achieve sub‑100 ms inference latency for generation workloads.
- Define and enforce rigorous evaluation metrics (semantic search: nDCG, Recall@K, MRR; generation: grounding accuracy, hallucination rate).
- Collaborate seamlessly with engineering, product, and research teams.
**Key Responsibilities**
1. Design & fine‑tune LLMs for multimodal generation tasks.
2. Build, productionize, and maintain RAG pipelines integrating embedding‑based search, metadata filtering, and LLM‑driven re‑ranking/summarization.
3. Apply prompt engineering, RAG techniques, and model distillation (LoRA/QLoRA, checkpoint reloading) to enhance grounding and reduce hallucinations.
4. Establish and automate evaluation frameworks for ranking, retrieval, and generation quality.
5. Optimize inference pipelines (token budgeting, prompt compression) to meet latency targets.
6. Deploy models at scale using distributed inference setups.
7. Work with product and research teams to prototype multimodal integrations for user-facing applications.
**Required Skills**
- Deep knowledge of NLP, machine learning, and multimodal AI.
- Hands‑on experience with LLM fine‑tuning, RAG, distillation, and large‑scale deployment.
- Proficiency in semantic search frameworks (FAISS, Weaviate, Vespa, Pinecone).
- Strong understanding of evaluation metrics for ranking (nDCG, Recall@K, MRR) and generation (grounding accuracy, hallucination rate).
- Expertise in Python, PyTorch/TensorFlow, and modern ML toolkits.
- Proven track record shipping latency‑sensitive AI products.
- Excellent communication and collaborative skills across global teams.
**Required Education & Certifications**
- Bachelor’s or Master’s degree in Computer Science, Electrical Engineering, Statistics, Machine Learning, or related field.
- Advanced coursework or experience in NLP, deep learning, and large‑scale ML systems.
- Industry certifications (e.g., TensorFlow, PyTorch, AWS ML) are a plus but not mandatory.
Los angeles, United states
On site
Senior
26-12-2025