Pavan Wale
Senior Machine Learning Engineer @BT Group
Signup · Get unlimited contacts
WORK HISTORY
Senior Machine Learning Engineer @BT Group
Bristol, GB
I have experience building and deploying AI-driven solutions, with a focus on LLMs and RAG. My projects include:Visual-LLM for Drone Image Analysis: Developed a VLLM system using AWS Bedrock and Anthropic\'s Claude-3 to automatically analyze drone images of towers, identifying and summarizing conditions like debris, contamination, rust, and deterioration. A Gradio interface allows users to easily upload images and receive reports.Generative AI PoCs:Nvidia NIM RAG Agent: Built a RAG agent using Nvidia NIM, LLaMa 3, and Langchain, incorporating adaptive, corrective, and self-RAG techniques for intelligent question answering. It uses Nvidia NIM models, ChromaDB, and Tavily Search.Wolfram Mathematica Integration: Created a Langgraph agent that seamlessly integrates Wolfram Mathematica and AWS Bedrock, routing computational questions to Wolfram and other questions to the Bedrock LLM.LLMOps for Contract Management: Developed a system using DocAI, Gemini, and GCP services (GCS, BigQuery, Cloud SQL) to extract information from PDF contracts, generate embeddings, and provide user-friendly responses to contract-related queries.ML Deployment Templates: Created templates to streamline ML model deployment on GCP, reducing the process to 14 days. These templates have been used to successfully deploy six use cases, leveraging various GCP services and MLOps best practices.Automated GCP Project Creation: Led the development of an automated system for creating GCP projects, improving efficiency for data science teams. This solution uses Python, Terraform, and other GCP services.
EDUCATION
KLS Gogte Institute of Technology
Bachelor’s Degree, Computer Science
SKILLS
ABOUT PAVAN WALE
Senior AI/ML Engineer with 9+ years of industry experience and specializing in LLMs, MLOps and Data Science, developing intelligent solutions for diverse applications. Key projects include a Visual-LLM for drone image analysis using AWS Bedrock, a RAG agent leveraging Nvidia NIM and LLaMa 3 for enhanced question answering, and a hybrid agent integrating Wolfram Mathematica and Bedrock. Built a RAG-based contract management system on GCP, automating data extraction and response generation. Streamlined ML deployments, reducing time to 14 days, and automated GCP project creation.Proficient in Python, LLMs, Langchain, Nvidia NIM, GCP, AWS, and MLOps. Additional experience in data pipeline development, infrastructure automation, data analysis, and instruction.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.