O assunto discutido aqui é sobre o desenvolvimento e a operacionalização de um pipeline robusto de Data Science focado na detecção não supervisionada de anomalias para enfrentar o risco de desvios e fraudes que se amplificou com a escalada do volume e da complexidade das ordens de serviço (OS) em garantia na indústria de linha branca. O objetivo central deste trabalho foi ranquear as ordens de serviço abertas pelo seu grau de anomalia, uma tarefa que exigiu a integração e higienização de cinco bases de dados corporativas e uma engenharia de atributos (features) comportamentais e contextuais extremamente detalhada, para capturar desvios sutis que se desviem do padrão histórico "normal". O algoritmo Isolation Forest (IForest) foi treinado em um espaço de 24 Componentes Principais (PCs), gerando um escore contínuo de anomalia para cada caso, o que fornece um mecanismo robusto para priorizar auditorias no pós-vendas, aumentar a eficiência da triagem de casos críticos e, assim, mitigar perdas e riscos operacionais.

INOV 006 | DataScience - Detecção de anomalias no pós-vendas da industria

9.31 bats around the first round of the MLB ‘25 playoffs (0:43), walks through a quiz on 2025 MLB full season (5:27), reviews Hunter Brown (21:53), and reviews Jose Altuve (28:19).

Episde 9.31 - Easy Grader

9.32 bats around MLB ‘25 playoff divisional round (0:40), continues cracking into curveball data (6:37), reviews Michael Busch (22:10), and reviews Matt Boyd (30:19).

Episode 9.32 - Dennis Green Player

Dr. Bhramar Mukherjee, Anna M.R. Lauder Professor of Biostatistics and Professor of Chronic Disease Epidemiology at the Yale School of Public Health, explores her research on “Shaping the Future of Public Health Data Science: Leadership, Mentorship, and Impact,” offering insights that both inspire and inform.

This episode features a personal and professional conversation about Professor Mukherjee’s leadership journey from academia to global recognition and her pivotal role in training the next generation of public health data scientists through NIH-funded programs.

Key discussion points include:

• Founding and leading a data science training program for undergraduates
• Building inclusive pathways in STEM and public health
• Reflections on awards, leadership roles, and driving systemic change
• Key highlights from the India edition of the Harvard Data Science Review

This podcast is brought to you by Global Management Consultancy. For more information, please visit www.globalmanagementconsultancy.com.

Disclaimer:

A. The background music used in this video is the property of its respective developer and is protected by Copyright. Although it is a free version, Business Talk, Global Management Consultancy and Deepak Bhatt do not hold the rights to this music.

B. Dr. Bhramar Mukherjee has generously shared profound insights from her research, “Shaping the Future of Public Health Data Science: Leadership, Mentorship, and Impact,” during her appearance on the Business Talk podcast channel. The uploaded video contains copyrighted material; therefore, any modifications to graphics, music, or the presence of the author or host are strictly prohibited.

Public Health Data Science Insights from Yale’s Dr. Bhramar Mukherjee

9.33 bats around some MLB ‘25 Championship series (0:41), rides some Slow Horses (3:52), works through some descriptive statistics on Runs (5:55), reviews Nathan Lukes (19:44), and reviews Shane Bieber (26:46).

Episode 9.33 - Feelable Numbers

9.34 bats around some WS observations before game 6 (0:38), notes the Twins manager search (5:25), looks through the MLB prediction machine results (6:28), reviews Will Smith (19:07), reviews Tyler Glasnow (26:35).

Episode 9.34 - Middle Bins

Efficient data pipelines are the foundation of every successful AI project, ensuring clean, accurate, and actionable insights. Destinie James highlights how strong data infrastructure empowers innovation, scalability, and smarter decision-making across AI-driven organizations worldwide. Visit here: https://destiniejames.journoportfolio.com/


Destinie James Talks About How Data Pipelines Drive AI Success

9.35 rates the World Series (0:42), explores saves (5:31), reviews Noelvi Marte (20:11), and reviews Hunter Greene (26:50).

Episode 9.35 - Within Margin of Error

Tonight, Thursday, November 13th, at 6 pm on The Brian Crombie Hour I interview ✅ Eric Guntermann. 👉 ALL my 🎤 podcasts and 🎦 video casts can be accessed any time on my web site https://briancrombie.com which holds repository of all my shows. 📌 Eric Guntermann, PhD, is a founder and chief data scientist at Empirical Intelligence, a Toronto-based research firm redefining how we understand public opinion. We explore how he entered polling, what Canadians really think of Prime Minister Mark Carney, and how voter loyalties are shifting across parties and generations. Eric explains why polls sometimes miss the mark, how U.S.–Canada tensions shape opinion, and what his latest data reveal about the national mood. He also breaks down conjoint analysis, the new gold standard for decoding real-world choices.

📻 The Brian Crombie Hour airs nightly at 6 PM on Sauga960AM 
🎧 Podcasts: https://briancrombie.com/podcasts-the-brian-crombie-hour 
📺 Videos: https://briancrombie.com/video-gallery 
📌 Subscribe to Brian’s YouTube Channel: https://bit.ly/2Oioec8 

#BrianCrombie #CanadaPolitics #PublicOpinion #Polling #Elections #DataScience #Leadership #MarkCarney #EricGuntermann

Eric Guntermann | Polling

9.36 wades through MLB awards voting (0:40), scrutinizes 2025 preseason predictions (5:17), reviews Geraldo Perdomo (19:35), and reviews Brandon Pfaadt (27:39).

Episode 9.36 - Obvious Trend

Destinie James is an accomplished Technical Program Manager, engineer, and global STEM advocate whose work bridges technology, leadership, and impact. She holds multiple engineering degrees in computer and electrical engineering and has blended her expertise in artificial intelligence, cybersecurity, and digital transformation to create meaningful change across industries.

To quote Dr. Evil, "The details of our lives are quite inconsequential." What's important is that we like fantasy, as it applies to the genre and the game derived from baseball statistics. We also enjoy developing analytical and computational “tools” for supporting the latter habit. Our goal is to share our fantasy baseball tools with a little help from the fantasy genre for context, allusion, and (fire and) ice breaking. In doing so, we hope to broaden the way our audience approaches fantasy baseball. Our podcast presents our methods to the public in the spirit of OpenSource - through freely helping others we promote active conversation that improves our tools.

Welcome to Business Talk, your go-to podcast for the latest trends, insights, and thought-provoking discussions in the business world. Whether you're a business professional, entrepreneur, researcher, or academic, our episodes will challenge you to rethink conventional wisdom and inspire actionable ideas.

Brought to you by Global Management Consultancy, we are committed to driving innovation and excellence in the business community. All content Copyrighted 2024 by Global Management Consultancy.

For more information about our past and upcoming podcasts, please click here:https://www.deepakbbhatt.com/businesstalk

Além de especialista em OKR e BSC é gestor de TI e cientista de dados, com mais de 14 anos conectando tecnologia a planejamento estratégico, indicadores e resultados. Lidera times multidisciplinares (Infra, BI, SAP, RPA e Processos), estruturando Data Lake, pipelines de BI e KPIs executivos para decisão ágil. É criador da plataforma PlanningBI, que integra OKR + BSC para alinhar metas, executar com governança e medir impacto. Ao longo da carreira, automatizou rotinas críticas, reduziu horas manuais e elevou a performance operacional e financeira, sempre com foco em previsibilidade e ROI.

Brian Crombie has 20 years of corporate experience, a depth of knowledge in all facets of finance, strategic planning and corporate / business development for a wide breadth of companies in several industries. Brian has deep political, community and charitable involvement as well.

New tracks tagged #datascience