[Webinar] Top KPIs for Data-Driven Quality Management

Interview with Khanh Pham, Head of Data Science at Inspectorio
Resiliency and reputation are extremely important in today’s commerce environment — and both depend on your ability to maintain product quality regardless of the size and complexity of your supply chain.
As the person charged with ensuring product quality, managing a global supply chain of 1000+ vendors and factories presents a challenge. And when addressing product quality issues, conventional quality control methods can quickly turn a precarious situation into full-blown firefighting.
Yet without a way to view the real-time quality risk of every factory in your supply chain, it’s not possible to truly know where the risk lies.
Any attempt to prevent such a situation will require significant surveillance, reporting, and expenditure and can take several weeks to complete — by which time the damage is done.
To avoid this, your team needs to work with data science to prevent quality defects from the root.
Ensuring quality across a large number of suppliers requires a high level of visibility, metrics, and efficiency. Data science with a digital network platform provides you with all three on an unprecedented scale.
Suddenly the task of analyzing risk is delegated to robust AI and machine learning software, providing you with 24/7, real-time, wholly autonomous quality risk monitoring.
Strategy and budget become clearer and more predictable. You can accomplish better quality control than ever before with less money and time.
Below we speak with Khanh Pham, Head of Data Science at Inspectorio, to understand how he helps Inspectorio’s clients automate their quality programs and mitigate risk — dramatically improving operational efficiencies and product quality.
Learning math was always “like solving a puzzle” to Khanh. From early childhood through his university career and Ph.D. in Computer Engineering, mathematics have been one of Khanh’s greatest strengths.
Years ago, he brought his capabilities to Inspectorio, leading the development of Inspectorio’s machine learning technology.
The benefits of technology and machine learning are penetrating the quality control industry, and Khanh has been at the forefront of helping the biggest brands and retailers in the world build an autonomous risk-based quality program.
Khanh now explains the impact data science and machine learning are having and will continue to have in the quality control industry.
Khanh: “Data science is both an art and a science. We use it to extract meaning from data to provide valuable insights that help people make better decisions.
We provide three layers of insights: descriptive (which tells what happened in the past), predictive (which learns from past patterns to forecast what may happen in the future), and prescriptive (which provides advice and recommendations).”
Khanh: “The goal of Inspectorio is to provide an autonomous system that makes quality management more efficient, effective, and accurate — the data science team is at the forefront of using machine learning to accomplish this goal for our clients, which include the largest brands and retailers in the world, as well as their suppliers.”
Khanh: “Our Factory Risk Prediction and Defect Recommender are good examples of how we help brands and retailers improve product quality with the latest technologies.
Let’s begin explaining the Factory Risk Prediction. There are several factors that make up a statistical model of factory risk — inspection fail rate, defect fail rate, fail rate over time, etc. We consider a lot of different factors to develop a comprehensive picture of factory risk like no other technology company in the industry.
Our Factory Risk Prediction model is dynamic and can change over time with each new inspection, because factory performance can change over time. Each piece of incoming data is used in our model to predict factory risk for a more objective and data-driven pattern.
This allows our clients to have real-time visibility about the performance of their factories, and ultimately make up-to-date business decisions to keep up with the industry and the competition.
Next, we have our Defect Recommender, which allows us to learn from the historical data of the factory. We assess what defects have occurred in the past for similar factories and similar products. As each new inspection comes, we can use that historical data to recommend what future defects might occur. This gives our clients insights into where to look for potential issues and proactively mitigate risk.
We’re also developing a Product Risk Prediction, which takes into account risk based on a comprehensive set of inputs from the historical performance of similar products, such as lab test results, raw materials, product description, and complexity. This is the next step to keep building on our automations and risk prediction. We’re working towards this vision to better serve our clients in their journey to full supply chain automation.”
Khanh: “The first is better decision making. With clearer assessment of factories and products, brands and retailers are optimizing the efficiency of their quality management programs. They can focus their attention and resources on high-risk areas to prevent defects and increase the performance of the factories.
Secondly, when they have the correct assessment of low-risk areas, they can reduce the resources consumed by those areas. This helps save cost and make monitoring more effective and efficient.”
Khanh:
“The main vision of our company is to build a collaborative performance platform for quality and compliance. We’re building a really unique end-to-end performance management platform that helps brands, retailers, and suppliers perform better and collaborate better.
We are able to guide our users regarding the kinds of actions they should take and the areas they should pay attention to. The benefits are immense and many of our clients are already experiencing them.”
Khanh: “Remember that it’s a journey. Getting from the visualization process, to the descriptive level, to the predictive level, to the prescriptive level takes time. If that’s their vision, they need to remember that it’s step-by-step.
However, this transformation is a good change — and a journey that brings many rewards in the short and long-term. It’s where the whole industry is going, and brands and retailers need it to maintain a competitive advantage.”
Pham: “If they want to standardize and optimize their quality operations with machine learning and automation, they should work with technology partners that have the appropriate data science expertise to deliver on the benefits this kind of technology can bring.
Like I mentioned previously, enabling an automated risk-based quality program is a journey — brands and retailers need to choose a partner that will work with them throughout the entire journey.”
At Inspectorio, our products use machine learning to create automated quality programs for the world’s largest brands and retailers. If your company has a global supply chain, Inspectorio can help you:
About the interviewee
Khanh Pham
Head of Data Science at Inspectorio
LinkedIn