BEGIN:VCALENDAR VERSION:2.0 X-WR-CALNAME:EventsCalendar PRODID:-//hacksw/handcal//NONSGML v1.0//EN CALSCALE:GREGORIAN BEGIN:VTIMEZONE TZID:America/New_York LAST-MODIFIED:20240422T053451Z TZURL:https://www.tzurl.org/zoneinfo-outlook/America/New_York X-LIC-LOCATION:America/New_York BEGIN:DAYLIGHT TZNAME:EDT TZOFFSETFROM:-0500 TZOFFSETTO:-0400 DTSTART:19700308T020000 RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU END:DAYLIGHT BEGIN:STANDARD TZNAME:EST TZOFFSETFROM:-0400 TZOFFSETTO:-0500 DTSTART:19701101T020000 RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU END:STANDARD END:VTIMEZONE BEGIN:VEVENT CATEGORIES:College of Engineering,Thesis/Dissertations DESCRIPTION:Faculty Supervisor: Dr. Amir Akhavan Masoumi, Computer & Inform ation Science/Data Science Committee Members: Dr. AshokKumar Patel, Comput er & Information Science/Data Science Dr. Debarun Das, Computer & Informat ion Science/Data ScienceÌýLocation/Link: Online via Zoomhttps://us04web.zo om.us/j/71177187003?pwd=bfh7typ8TW4oqb7tPqGZ7GMqY6Zpa7.1 Meeting ID: 71177 187003Passcode: tt8zdaÌýAbstract:The rapid growth of online retail has cre ated enormous volumes of unstructured product data that most businesses st ruggle to turn into actionable intelligence. This study presents an intell igent analytics platform that combines Retrieval-Augmented Generation (RAG ) with Claude Opus 4.6 to generate structured business insights from a cor pus of 200,000 Amazon Electronics product records. A multi-layered pipelin e transforms raw product metadata into semantically rich text chunks, enco des them using BGE-M3 sentence embeddings, and stores the resulting 200,00 0 vectors in a ChromaDB persistent vector store. At query time, the platfo rm retrieves the most contextually relevant product records, reranks them by semantic similarity, and feeds them to Claude Opus 4.6, which synthesiz es the retrieved evidence into coherent, data-grounded analytical narrativ es complete with business recommendations. The platform is built with prod uction deployment in mind, with MLflow tracking every experiment for full reproducibility, Docker containerizing the entire application stack, and G itHub Actions automating the continuous integration and delivery pipeline. An interactive Streamlit dashboard brings all capabilities together in a user-friendly interface requiring no technical expertise. Evaluation acros s eight quantitative metrics confirms the quality of the system's outputs, achieving a ROUGE-1 score of 0.4121, a ROUGE-L score of 0.4121, and a BER TScore F1 of 0.9131, indicating strong lexical precision and exceptional s emantic alignment with human-authored reference insights. A faithfulness s core of 0.5567 demonstrates that generated content is reliably grounded in retrieved evidence. All sixteen automated unit tests pass, confirming the robustness of every system component.ÌýFor further information, please co ntact Dr. Amir Akhavan Masoumi at aakhavanmasoumi@umassd.edu.\nEvent page: /events/cms/rag-powered-customer-insight-generation -for-e-commerce-using-llms-vector-search-and-an-end-to-end-mlops-pipeline. php\nEvent link: https://us04web.zoom.us/j/71177187003?pwd=bfh7typ8TW4oqb7 tPqGZ7GMqY6Zpa7.1 X-ALT-DESC;FMTTYPE=text/html:

91É«ÇéƬ

Faculty Supervisor: Dr. Amir Ak havan Masoumi\, Computer & Information Science/Data Science

\n

Commit tee Members:

\n

Dr. AshokKumar Patel\, Computer & Information Science /Data Science

\n

Dr. Debarun Das\, Computer & Information Science/Dat a Science
Ìý
Location/Link: Online via Zoom

\n

Meeting ID: 71177187003
Passcode: tt8zda
Ìý
Abstra ct:
The rapid growth of online retail has created enormous volumes of unstructured product data that most businesses struggle to turn into acti onable intelligence. This study presents an intelligent analytics platform that combines Retrieval-Augmented Generation (RAG) with Claude Opus 4.6 t o generate structured business insights from a corpus of 200\,000 Amazon E lectronics product records. A multi-layered pipeline transforms raw produc t metadata into semantically rich text chunks\, encodes them using BGE-M3 sentence embeddings\, and stores the resulting 200\,000 vectors in a Chrom aDB persistent vector store. At query time\, the platform retrieves the mo st contextually relevant product records\, reranks them by semantic simila rity\, and feeds them to Claude Opus 4.6\, which synthesizes the retrieved evidence into coherent\, data-grounded analytical narratives complete wit h business recommendations. The platform is built with production deployme nt in mind\, with MLflow tracking every experiment for full reproducibilit y\, Docker containerizing the entire application stack\, and GitHub Action s automating the continuous integration and delivery pipeline. An interact ive Streamlit dashboard brings all capabilities together in a user-friendl y interface requiring no technical expertise. Evaluation across eight quan titative metrics confirms the quality of the system's outputs\, achieving a ROUGE-1 score of 0.4121\, a ROUGE-L score of 0.4121\, and a BERTScore F1 of 0.9131\, indicating strong lexical precision and exceptional semantic alignment with human-authored reference insights. A faithfulness score of 0.5567 demonstrates that generated content is reliably grounded in retriev ed evidence. All sixteen automated unit tests pass\, confirming the robust ness of every system component.
Ìý
For further information\, ple ase contact Dr. Amir Akhavan Masoumi at aakhavanmasoumi@umassd.edu.

Event page: /events/cms/rag-powered-cus tomer-insight-generation-for-e-commerce-using-llms-vector-search-and-an-en d-to-end-mlops-pipeline.php
Event link:

< /html> DTSTAMP:20260519T072520 DTSTART;TZID=America/New_York:20260518T143000 DTEND;TZID=America/New_York:20260518T153000 LOCATION:Online - Zoom SUMMARY;LANGUAGE=en-us:RAG-Powered Customer Insight Generation for E-Commer ce Using LLMs, Vector Search, and an End-to-End MLOps Pipeline UID:9d1b7a4adf14a16a781391e714bf271f@www.umassd.edu END:VEVENT END:VCALENDAR