Parul Sharma

AI/ML Engineer – Financial AI & Machine Learning

• Designed and deployed an LLM-powered financial assistant using LangChain and OpenAI GPT-4, enabling natural-language portfolio analysis and achieving 92% intent-recognition accuracy
• Implemented a Retrieval-Augmented Generation (RAG) architecture using Pinecone for semantic document retrieval across 10,000+ financial reports, reducing research time by 60%
• Built a domain-specific financial sentiment analysis pipeline using BERT and LoRA fine-tuning, achieving an F1 score of 0.89 for market sentiment classification
• Automated trade-recommendation and risk-assessment workflows by developing conversational AI agents with LangChain and function calling
• Established prompt-engineering and LLM evaluation framework with systematic testing and performance metrics, enhancing response quality by 35%
• Engineered LSTM-based time-series forecasting models for stock-trend prediction, achieving an 18% improvement in MAE compared with baseline models
• Developed a collaborative-filtering recommendation engine using matrix factorization for investment-asset recommendations, increasing user engagement by 25%
• Engineered end-to-end MLOps pipeline with MLflow, Docker, and Kubernetes for automation of model training, experiment tracking, model versioning, and deployment

• Developed and deployed machine learning models using Python and Scikit-learn to solve business problems.
• Applied hyperparameter tuning to improve model performance and achieve 60% improvement in accuracy/F1-score.
• Performed data preprocessing, feature engineering, exploratory data analysis (EDA), and model evaluation.
• Built NLP models for text classification/sentiment analysis using relevant NLP techniques.
• Created interactive dashboards with React and D3.js visualizing ML model predictions and sensor data streams
• Implemented REST APIs using Node.js and FastAPI for serving ML predictions with sub-100ms latency
• Deployed microservices architecture on AWS ECS and Azure Kubernetes Service for scalable ML inference
• Integrated Azure IoT Hub and AWS IoT Core for real-time data ingestion and processing pipelines