Skip to content
/ fmit Public

Repo for Failure Mode Identification Tool using Large Language Models and Clustering for Databricks Hackathon

Notifications You must be signed in to change notification settings

msraihan/fmit

Repository files navigation

fmit

The Failure Model Identification Tool streamlines warranty claims processing by analyzing and summarizing detailed claim inputs, allowing users to gain quick insights into customer issues. Leveraging natural language processing (NLP), the chatbot reads through claim texts, extracts key information such as the customer complaint, probable cause, and the corrective action taken, and then provides a concise summary of each claim. Beyond summarization, the tool organizes data by clustering similar claims into groups of common failure modes, helping manufacturers identify recurring issues and potential product weaknesses. With this functionality, teams can more efficiently pinpoint problematic components, improve product reliability, and proactively address common failure patterns. Ideal for enhancing customer support, technical diagnosis, and product development, this analytics solution enables faster, data-driven decision-making across warranty management processes.

Summarized Table Generation.py - Notebook to create summarized data using large language models for both initial summarization and vector embedding of different relevant information Home.py - Main streamlit app file to run

About

Repo for Failure Mode Identification Tool using Large Language Models and Clustering for Databricks Hackathon

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages