Data Scientist

Osmii; Greater London; Hybrid; GBP 90,000 - 130,000; Full time

Job summary

Osmii is seeking a Data Scientist in London with GenAI expertise to lead the development and deployment of enterprise-grade GenAI solutions. You will work on LLM-based applications, prompt engineering, fine-tuning, embeddings, and retrieval-augmented generation (RAG).

Ideal candidates have strong ML/NLP foundations, hands-on GenAI framework experience, and cloud environment familiarity. This role supports flexible hybrid working in London and collaborates with cross-functional teams.

Qualifications

• 5+ years in Data Science / ML; 1+ year delivery in LLM / GenAI projects.
• Strong Python skills with Transformers, LangChain, scikit-learn, PyTorch or TensorFlow.
• Experience with foundation models (OpenAI GPT-4, Claude, Mistral, LLaMA).
• Semantic search and embeddings (vector search, BERT, Sentence Transformers).
• Deployment of AI workflows via APIs or web frameworks (FastAPI, Streamlit, Flask).
• Cloud & MLOps familiarity (AWS, GCP, Azure).
• Excellent communication; translate tech to business impact.
• Hands-on with prompt tuning, few-shot, or LoRA fine-tuning.
• Data privacy, security, governance in GenAI.
• Experience in enterprise/regulatory domains a plus (finance/healthcare/insurance).
• Palantir tools knowledge a bonus.

Responsibilities

• Design & Build GenAI solutions across enterprise use cases: customer service, document automation, summarization, retrieval.
• Fine-tune and adapt foundation models using domain datasets.
• Implement RAG pipelines, embedding models, vector databases (FAISS, Pinecone, ChromaDB).
• Collaborate with data engineers, MLOps, and product teams to build AI apps and APIs.
• Develop prompts and prompt chains using LangChain, LlamaIndex, PromptFlow, or custom code.
• Evaluate model performance, mitigate bias, optimize accuracy, latency, cost.
• Stay updated with LLMs and GenAI system architecture.

Skills

Python

GenAI/LLM

Prompt engineering

LangChain

Deployment via APIs

Cloud & MLOps

Communication

Few-shot / LoRA

Data privacy & governance

Enterprise AI

Transformers / PyTorch / TensorFlow

Tools

Transformers

LangChain

scikit-learn

PyTorch

TensorFlow

FAISS

Pinecone

ChromaDB

FastAPI

Streamlit

Job description

London, United Kingdom (Flexible hybrid working)

Permanent

ROLE SUMMARY

We are seeking a highly capable and innovative Data Scientist with specialized experience in Generative AI to join our Data Science Team. In this role, you will lead the development and deployment of enterprise-grade GenAI solutions, including LLM-based applications, prompt engineering, fine-tuning, embeddings, and retrieval-augmented generation (RAG).

The ideal candidate possesses a strong foundation in machine learning and NLP, paired with hands-on experience using modern GenAI frameworks and cloud environments.

KEY RESPONSIBILITIES

• Design & Build: Develop Generative AI solutions using Large Language Models (LLMs) across enterprise use cases such as customer service, document automation, summarization, and knowledge retrieval.
• Model Adaptation: Fine-tune and adapt foundation models using domain-specific datasets.
• Pipeline Architecture: Implement RAG pipelines, embedding models, and vector databases (e.g., FAISS, Pinecone, ChromaDB).
• Cross-functional Collaboration: Partner with data engineers, MLOps, and product teams to build end-to-end AI applications and APIs.
• Prompt Engineering: Develop custom prompts and prompt chains utilizing frameworks like LangChain, LlamaIndex, PromptFlow, or custom code.
• Optimization & Governance: Evaluate model performance, mitigate bias, and optimize accuracy, latency, and operational cost.
• Innovation: Stay current with the latest advancements in LLMs, transformers, and GenAI system architecture.

Essential Qualifications

• 5+ years of professional experience in Data Science / ML, with 1+ year of hands-on delivery in LLM / GenAI projects.
• Strong Python Skills: Expertise with libraries such as Transformers, LangChain, scikit-learn, PyTorch, or TensorFlow.
• LLM Expertise: Direct experience working with foundation models such as OpenAI (GPT-4), Claude, Mistral, LLaMA, or similar architectures.
• Semantic Search & Embeddings: Deep understanding of vector search, embedding models (e.g., BERT, Sentence Transformers), and retrieval techniques.
• Deployment: Demonstrated ability to build scalable AI workflows and deploy them via APIs or web frameworks (e.g., FastAPI, Streamlit, Flask).
• Cloud & MLOps: Familiarity with major cloud providers (AWS, GCP, Azure) and MLOps best practices.
• Communication: Exceptional ability to translate complex technical solutions into strategic business impact.
• Hands-on experience with prompt tuning, few-shot learning, or LoRA-based fine-tuning.
• Practical knowledge of data privacy, security, and governance considerations within GenAI systems.
• Familiarity with enterprise architecture, SDLC, or building solutions within highly regulated domains (finance, healthcare, insurance).
• Knowledge of Palantir tools is a strong bonus