Data Scientist - Applied AI Scientist

Responsibilities

• End-to-End AI Design: Design, prototype, and validate ML/AI solutions, translating complex business challenges into mathematical formulations and scalable, production-ready code.
• Advanced Analytics & EDA: Perform deep exploratory data analysis, statistical testing, and data transformations on diverse datasets (structured and unstructured) to uncover predictive signals and validate hypotheses.
• Production-Grade Science: Architect and implement modular, extensible, and testable Python codebases for AI experiments. Move beyond Jupyter notebooks by applying clean-code principles (SOLID, DRY) for seamless hand-off to ETO Engineering teams.
• Agentic & Generative AI: Develop and experiment with applied generative AI and multi-agent architectures using orchestration frameworks (e.g., LangChain, LangGraph), focusing on optimal state management, robust RAG pipelines, and efficient system design.
• Algorithmic Optimization: Optimize model inference, data processing pipelines, and memory footprints for latency and scalability, applying a strong understanding of data structures and algorithmic complexity.
• Rigorous Evaluation: Build automated evaluation frameworks to benchmark model performance, mitigate hallucinations, track drift, and ensure algorithmic fairness via A/B testing and statistical rigor.
• Collaboration & Communication: Act as the technical translator between research-focused ideation and engineering execution. Communicate complex statistical findings and system architectures to both technical and non-technical stakeholders.

Qualifications

• The Scientist's Mind: Solid foundation in statistics (Bayesian/Frequentist), linear algebra, hypothesis testing, and the internal mechanics of ML algorithms (e.g., how optimizers work, loss functions, attention mechanisms).
• The Engineer's Toolbelt: Advanced Python proficiency with a strong focus on Object-Oriented Programming (OOP) and modular design. You must be comfortable writing unit tests (e.g., Pytest) for your data pipelines and models.
• Framework Depth: Deep expertise with ML libraries (PyTorch, TensorFlow, Scikit-learn, Pandas) and experience implementing custom logic, rather than just calling out-of-the-box models.
• Generative AI Systems: Hands-on experience with NLP, Large Language Models (LLMs), and Vector Databases, with an understanding of how to evaluate and optimize these systems at scale.
• Software Maturity: Proficiency with Git/version control, containerization (Docker), API development (FastAPI/Flask), and a working knowledge of how models fit into a CI/CD lifecycle (MLOps).
• Problem Solving: Exceptional problem-solving skills, comfort with ambiguity, and the ability to own the data science lifecycle from abstract ideation to engineered prototype.