Poonam Deshmukh

• Developed and deployed Generative AI solutions using LLMs and RAG pipelines, enhancing documentation accuracy by 12% and increasing model precision by 33%.

Project Portfolio

AI Driven Assistant for Cosmetic Dermatology Procedures.

• Developed Generative AI models to automate quality assessment of dermatology procedures.
• Implemented RAG, LLMs and transformers for quality insights and automated validation.
• Improved precision by 33% via fine-tuning of retriever-generator components.

• Deployed ML models, improving operational decision-making and reducing manual effort by 30%.
• Implemented AI solutions that boosted project success rate by 25% and improved customer engagement insights by 20%.

Project Portfolio

Crop Yield Prediction using Soil Analysis and Crop Recommendation.

• Developed machine learning models to predict crop yield based on soil parameters and climate data.
• Applied feature engineering and data preprocessing techniques to enhance model accuracy
• Utilized Random Forest and regression models for predictive analysis
• Integrated AI recommendations for precision agriculture. Improved crop yield prediction accuracy by 18%.
• AI-driven recommendations for precision farming and resource optimization.
• Collaborated with agricultural experts to validate model predictions and improve decision-making

Skin Sensitivity Prediction & Sentiment Analysis.

• Predicted skin sensitivity with 87% accuracy using user/ingredient data.
• Conducted NLP-based sentiment analysis with 85% accuracy to guide product strategy