Data Scientist - C2H Role

Role & responsibilities

• Job Description:

• As a Senior Data Scientist specializing in Fraud and Anomaly Detection

  ,
• you will play a pivotal role in developing and implementing advanced models and algorithms to detect fraudulent activities and anomalies within complex datasets.
• You will leverage your expertise in Graph Anomaly Detection and Graph Neural Networks to enhance our anomaly detection capabilities.
• Key Responsibilities: o Experience: o Candidate must have hands-on experience in the field of Anomaly/Fraud/Outlier Detection and the connected technologies.
• Advanced Model Development and Implementation: o Design and develop sophisticated machine learning models specifically tailored for fraud and anomaly detection using state-of-the-art techniques.
• Implement Graph Neural Networks (GNNs) using PyTorch Geometric to capture complex relationships and dependencies in graph-structured data. o Develop algorithms for Graph Anomaly Detection, focusing on node, edge,
• and subgraph anomaly identification. o Have a know-how of various Graph Neural Network algorithms like GCN, GraphSAGE, GAT, GINE, Graph Message Passing, Hetero Graph Learning etc. o Have very good understanding and know how of various NLP techniques and models like BERT,
• Text Classification, Text Extraction, LLMs and LLM Fine-Tuning etc.
• Data Analysis and Feature Engineering: o Perform in-depth exploratory data analysis (EDA) to uncover hidden patterns and insights related to fraudulent activities. o Engineer features from large-scale datasets, including constructing graph-based features such as node embeddings, edge attributes, and graph metrics.
• Utilize techniques such as spectral clustering and community detection to enhance feature representation in graph data. o Utilize various nlp techniques for data cleaning, visualization, data pre-processing and data preparation o Research and Innovation: o Conduct cutting-edge research on emerging methodologies in fraud detection, anomaly detection, and graph-based machine learning. INTERNAL o Experiment with novel architectures and algorithms, such as attention mechanisms in GNNs, to improve detection capabilities. o Prototype and evaluate new approaches using rigorous experimental design and statistical validation. o Collaboration and Communication: o Work closely with other data scientists to ensure seamless integration of models into production environments, optimizing for scalability and performance. o Collaborate with software developers to implement efficient data pipelines and real-time processing systems. Technical Skills: o Programming and Libraries o Proficiency in Python, with extensive experience in libraries such as PyTorch, PyTorch Geometric, TensorFlow, Keras, Transformers and Scikit-learn. o Candidate must know how to work with Graph Based Machine Learning and NLP. o Experience with graph processing frameworks and libraries, such as NetworkX and DGL (Deep Graph Library). o Machine Learning and Deep Learning: o Strong understanding of machine learning algorithms, including supervised, unsupervised, and semi-supervised learning techniques. o Expertise in deep learning architectures, particularly those applicable to graph data, such as Graph Neural Networks like GAT, GraphSAGE etc

Preferred candidate profile