Senior Data Scientist | CARE

View full description

Requirements:

• Master’s degree or PhD in Computer Science, Data Science, Machine Learning, Statistics, or a related field.
• Proficiency in Python and experience with machine learning frameworks such as PyTorch, TensorFlow, or similar.
• Strong understanding of generative AI architectures, including transformers and attention mechanisms.
• Experience in multi-agent systems and LLMs.
• Strong problem-solving abilities, with a focus on experimental design and data analysis.
• Excellent verbal and written communication skills, with the ability to explain complex technical concepts to non-technical stakeholders.
• Have clear scientific thinking and a passion for integration R&D and cutting-edge technology into a product
• Prior experience in Python and SQL

Responsibilities:

• Model Development & Fine tuning: Understanding of state-of-the-art deep learning techniques, such as Transformers architectures and attention mechanisms, and knowledge of fine-tuning LLMs for enhanced model performance, using techniques such as LoRA, QLoRA, among others.
• Function Calling & Tool Use: Design and implement structured function calling mechanisms within LLMs to enable dynamic interactions with APIs, databases, and retrieval-augmented generation (RAG) pipelines.
• Prompt Engineering: Craft clear instructions that help our models understand exactly what we need, creating reusable templates and testing against edge cases.
• Data Preparation: Preprocess, analyse, and curate high-quality datasets to train and fine-tune both embedding and generative models.
• Evaluation and Testing: Develop robust evaluation frameworks to assess agentic AI performance, combining automated evaluation (LLM-as-judge), adversarial testing, human-in-the-loop evaluations, and custom behavioral metrics.
• LLM Observability: Establish monitoring systems that track LLM behavior in production, capturing key metrics around information retrieval, hallucinations, and latency.
• Research and Innovation: Stay current with the latest research in generative AI,, share insights with the team, test promising approaches, and help make us better.
• Deployment: Collaborate with engineering teams to deploy models in scalable and efficient production environments.