subhash inavolu

Automated the data collection process, reducing manual workload and increasing data accuracy. Used ML model to identify the offer letter, from 40K PDFs from sharepoint Used Langgraph for Advanced RAG to extract data from supplier700+ offer letters, with an accuracy of 85%. The RAG model uses multistep context verification, initial reaction and reflection architecture inspired by DeepSeek Developed LLM Based AI model to compare documents - The output is a semanticaltechnical and commercial differences between versions- categorized by section Designed and implemented ML models for classifying various train assemblies in Dataiku,the newly improved feature engineering helped to build a model that improved the accuracy from 45 to 76% Built a web based user interface in Dash for users to interact with the Offer letter database Built statistical analysis model in Dataiku to find errors in BOM, by comparing data across different ERP systems Developed OOPS model in python to correct the weights of assemblies and individual components in PYTHON Developed experimental Multi agent orchestration in langgraph, with plan and execute architecture. The model answers questions on database, with query tool Developed advanced the use of AI-driven statistical models, to estimate the cost of assemblies, as a price distribution instead of a single values - Monte Carlo distribution in price estimate Built Power BI dashboards for should cost model, to negotiate with supplier and reference for cross-functional teams Won the award for Best idea in Development

Responsible for the Global Brake simulations softwore using matlab Upstream porject support for design and development of the brake system Predict the pedal feel, thermal performance of the brake system Analysis of the driving pattern by end user to optimize the brake system Analysis of the supplier test data to predict the brake performance Use Data engineering and machine learning to predict brake performance Development of new tools in python to analyse the brake pad life of different regions across the globe

Design and development of Brake system Vehicle testing of Commercial vehicles. Design of Brake on Ideas and unigraphics Calculation of Vehicle brake performance for Commercial Vehicles above 6 ton. Analysis of customer usage pattern