Continuous Product Quality Assurance (CPQA) Working Team

  Be a part of the conversation; join our    LinkedIn AI group    today!

Be a part of the conversation; join our LinkedIn AI group today!

Goal

To increase the assurance of our product quality by becoming a predictive organization through the power of AI to provide more robust information to make more informed decisions than we can today.
 

Deliverables

  1. System Map demonstrating interconnectivity and potential causal relationships between GMP systems, non-GMP systems and the complaint system:
    Medical Device System Map (PDF)
    Biopharma Data System Map (PDF)

  2. Summary of data elements that represent risks and/or potential failure modes for the events captured in the complaint system:
    Medical Device Data Element Model (Excel)
    Biopharma Data Element Model (Excel)

  3. Definitions of AI relevant terms to harmonize understanding.
    AI Dictionary (PDF)

Phase 2 (August 2018-August 2019)

Phase 2 Team Leadership

CO-LEADERS

  • Nick Ranly, Account Executive – Life Sciences, Siemens PLM Software

  • Mark Zhong, Associate, PwC Advisory Services

TEAM ADMINISTRATOR

  • Lacey Harbour, Regulatory Specialist - Validation Focus, Ken Block Consulting

MENTOR FROM AI CORE TEAM

  • Matthew Schmucki, Quality Lean Coach, AstraZeneca

AI EXPERT FROM AI CORE TEAM

  • Paul Kostoff, Managing Partner, Sphaeric.ai

Problem Statement

Despite the controls our industry has in place to assure product quality, we still have failures and recalls. The assurance of our product quality is limited by the ability of our organization to access, assess and connect all relevant data (known and unknown) real-time to inform our decisions.
 

Description of Work

  • Part 1: Develop a succinct business problem to solve that builds from the Phase I work accomplished on identifying the location of data across the TPLC linked to complaints received. Drive hard to interview non-quality and non-GXP peers to find out where their primary data sources are. Provide examples of how this non-GXP could influence product quality.

  • Part 2: Develop 2-3 different types of algorithms to demonstrate how the information can be scanned – including textual information (non-structured data) in documents. These algorithms can include linear regression versus decision tree modeling versus random forest, etc. Work to identify publicly available data to demonstrate the effectiveness of the algorithms.

  • Part 3: Develop a white paper to summarize to a lay-industry professional how AI is used. Walk through the processes from Parts 1 and 2, including the challenges that needed to be overcome, and how they were overcome.

Phase 1 Deliverables (August 2017-August 2018)

Phase 1 Team Leadership

CO-LEADERS

  • Charlene Banard, Senior Vice President – Global Quality, Shire

  • David Lonza, Head of EU Operations, Lachman

TEAM ADMINISTRATOR

  • Mark Zhong, Associate, PwC Advisory Services

MENTOR FROM AI CORE TEAM

  • Tony Barnes, Bucks County Advisors

AI EXPERT FROM AI CORE TEAM

  • Matthew Schmucki, Senior Validation Engineer, AstraZeneca

Problem Solving Process

  1. Map out the interconnected systems of the chosen system.

  2. Gather relevant variables/information (internal/external).

  3. Interview process of expert auditors and FDA investigators.