Parag Patil Email and Phone Number

Developed Q&A Chat Assistance using RAG, leveraging Azure Search and OpenAI Models. Integrated Knowledge Graph with Vector Databases like Redis into the existing RAG pipeline. Fine-Tuned GPT-4o and GPT-3.5 for downstream tasks to enhance the Master Data Management system (MDM). Custom-trained embedding models to enhance retrieval accuracy in Vector Databases. Utilized frameworks such as Langchain and Llamaindex for Generative AI solutions.

Engineered a recommendation engine for game and jackpot suggestions on slot machines, leading to a remarkable 40% improvement in client revenue.Conducted comprehensive analysis and discovery of AI-powered products, identifying opportunities for enhancement and optimization.Seamlessly integrated Generative AI capabilities with existing products, enhancing their functionality and appeal.Designed migration strategies and guidelines to facilitate smooth transitions for AI pipelines and products.Automated code creation for asset migration, including AI models and Python notebooks, streamlining the process and reducing manual effort.

Spearheading the development of AI/ML frameworks that cater to the unique needs of the Healthcare, EdTech, and HRTech sectors.Collaborating closely with VPs and Business Heads to align AI/ML strategies with industry demands, ensuring our solutions remain cutting-edge and relevant.Creating AI/ML accelerators that leverage NLP, Computer Vision, and advanced analytics to deliver innovative solutions that drive business growth.Partnering with a diverse range of clients to seamlessly integrate AI into their products and services, enhancing their competitiveness in the market.Keeping abreast of emerging techniques in the field, including NLP, GAN, MLOps, and AutoML, and applying this knowledge to our projects.Scaling AI/ML solutions efficiently through the use of MLOps, harnessing technologies such as Docker, AzureML, and Databricks to optimize performance and deployment.

Automated existing AI pipelines using Apache Airflow.Enhanced existing NLP models by incorporating state-of-the-art models such as GPT-3 and Roberta.

Proficiently executed Named Entity Recognition (NER) extraction from legal and contract documents using custom-trained models like Spacy and Flair, ensuring accuracy and efficiency in information extraction.Developed a deep learning NLP model for text summarization on legal documents, harnessing the power of the T5 transformer model to distill essential information effectively.Enhanced the performance of the NLP model "sec-vec" by transforming it into a more accurate text generation model "sec-sec" for precise clause extraction.Automated entity extraction and redaction in PDF documents, leveraging a combination of pre-trained and custom-trained models, streamlining data processing tasks.Built robust document classification models with advanced architectures like BERT and T5 (Text-To-Text Transfer Transformer), improving document organization and retrieval.Conducted web scraping of valuable information and documents from public government websites, contributing to data collection and analysis efforts.Developed a generic CAPTCHA breaker using Convolutional Neural Networks (CNN) and OpenCV, enhancing web data scraping capabilities by overcoming CAPTCHA challenges.Created efficient web services using Flask, Gunicorn, and Nginx, facilitating seamless data exchange and application deployment.Successfully deployed multiple ML projects in Quality Assurance (QA) and User Acceptance Testing (UAT) environments using Docker, ensuring project readiness for production deployment.

Utilized Python and AI/ML to automate manual processes.Developed and deployed NLP models for handling extensive fintech unstructured data.Built forecasting and survival analysis models to predict production server stability.Created analytical dashboards in real-time data using Power BI.