Looks like you have JavaScript disabled. For the full Twine experience, you will need to re-enable it.

AI/ML Engineer with 4+ years of experience building machine learning and generative AI solutions for finance, specializing in credit-risk and fraud detection. Experienced in end-to-end ML/MLOps using PyTorch, TensorFlow, Hugging Face Transformers, CatBoost, and LLaMA. Strong background in RAG-based pipelines, embeddings/vector databases (FAISS, Pinecone), and scalable deployments with Docker, Kubernetes, and MLflow. Experienced in real-time scoring, model monitoring, drift detection, and performance optimization across large-scale portfolio and fraud workflows.…AI/ML Engineer with 4+ years of experience building machine learning and generative AI solutions for finance, specializing in credit-risk and fraud detection. Experienced in end-to-end ML/MLOps using PyTorch, TensorFlow, Hugging Face Transformers, CatBoost, and LLaMA. Strong background in RAG-based pipelines, embeddings/vector databases (FAISS, Pinecone), and scalable deployments with Docker, Kubernetes, and MLflow. Experienced in real-time scoring, model monitoring, drift detection, and performance optimization across large-scale portfolio and fraud workflows.

Vaishavi Patel

AI Engineer, Data Scientist, Software Engineer, +2





AI/ML Engineer with 4+ years of experience building machine learning and generative AI solutions for finance, specializing in credit-risk and fraud detection. Experienced in end-to-end ML/MLOps using PyTorch, TensorFlow, Hugging Face Transformers, CatBoost, and LLaMA. Strong background in RAG-based pipelines, embeddings/vector databases (FAISS, Pinecone), and scalable deployments with Docker, Kubernetes, and MLflow. Experienced in real-time scoring, model monitoring, drift detection, and performance optimization across large-scale portfolio and fraud workflows.…AI/ML Engineer with 4+ years of experience building machine learning and generative AI solutions for finance, specializing in credit-risk and fraud detection. Experienced in end-to-end ML/MLOps using PyTorch, TensorFlow, Hugging Face Transformers, CatBoost, and LLaMA. Strong background in RAG-based pipelines, embeddings/vector databases (FAISS, Pinecone), and scalable deployments with Docker, Kubernetes, and MLflow. Experienced in real-time scoring, model monitoring, drift detection, and performance optimization across large-scale portfolio and fraud workflows.

Available to hire

AI/ML Engineer with 4+ years of experience building machine learning and generative AI solutions for finance, specializing in credit-risk and fraud detection. Experienced in end-to-end ML/MLOps using PyTorch, TensorFlow, Hugging Face Transformers, CatBoost, and LLaMA.

Strong background in RAG-based pipelines, embeddings/vector databases (FAISS, Pinecone), and scalable deployments with Docker, Kubernetes, and MLflow. Experienced in real-time scoring, model monitoring, drift detection, and performance optimization across large-scale portfolio and fraud workflows.

Skills

Experience Level

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Expert

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Intermediate

Beginner

Beginner

Beginner

Work Experience

AI/ML Engineer at Stripe

May 1, 2025 - Present

Built and deployed ensemble credit risk models using PyTorch and TensorFlow, achieving improvements in default-rate reduction and high-risk borrower identification. Designed and implemented a RAG pipeline using GPT-4 with Pinecone vector DB and dense embeddings to improve underwriting document retrieval accuracy. Engineered embedding-based borrower similarity index with BERT/Sentence Transformers to improve loss prediction precision and enable risk clustering. Implemented real-time scoring pipelines using Kafka, Spark, and TensorFlow Serving to reduce inference latency and support high-volume live scoring. Added automated model drift detection, performance monitoring, and retraining orchestration using MLflow, Airflow, and Great Expectations. Led A/B experiments using GPT/LLaMA-assisted scenario generation and risk forecasts, improving ROI and reducing loss provisioning expenses. Architected a centralized feature store with Feast to increase feature reuse and accelerate model delivery.

Machine Learning Engineer at Hexaware Technologies

January 1, 2021 - August 31, 2024

Constructed and deployed a multi-modal recommendation engine using graph neural networks and embeddings, improving CTR and average order value. Developed an NLP pipeline with Hugging Face Transformers for sentiment analysis and intent detection, achieving high F1 and enabling churn mitigation. Built a real-time fraud detection system leveraging XGBoost and LightGBM on streaming transaction data, reducing false positives and preventing annual losses. Implemented RAG-based knowledge retrieval with Pinecone and GPT/LLaMA embeddings to reduce support tickets and improve contextual reasoning. Built AutoML pipelines for tabular and time-series tasks, reducing training time and improving forecasting accuracy. Deployed end-to-end ML pipelines on AWS using Docker, Kubeflow, and MLflow for training/versioning/monitoring and real-time serving. Created explainability dashboards with SHAP and LIME and supported generative AI assistant experimentation for knowledge workflows.