Full Transcript
Access the complete word-for-word transcript for easy reference and accessibility.
All the steps are those blue dots. Those are the steps in your manufacturing process. And you can have five or you can have 55. It doesn’t matter. The way you basically keep adding those steps, they just keep showing up in the graph and then they keep getting connected, right? Now, in terms of the modalities and the flexibility, what this offers is it reflects your process. You don’t have to define a data model upfront.
Whatever you do from your manufacturing perspective reflects in the knowledge graph and that becomes your data model. That basically gives you a lot of flexibility. We had never tested cell therapy. We didn’t know whether it will work or it will not work. But the basic premise was that it should, as long as you’re doing the seven steps or the 18 steps, they will reflect and that did, right? And the thing is that this is true for the future as well.
When you bring in new modalities, your manufacturing process will be reflected on the knowledge graph and that becomes the way you organise your data. And if you organise your data, and this is basically all the data. It’s the fast moving data, it’s the temperature and the pressure and the flow rate. It’s the PAT, the Raman data that is coming in from your PAT equipment. It is your QMS data.
It is the data from your MES systems and LIMS and even from your ERP, no restriction. You have more systems there. It also allows you flexibility. If you don’t have any of the systems, all of the data is on paper, no problem. We have a specific way of basically bringing in the paper data and using it in the graph as well. With the business changes that are happening on a daily basis.
Today you’re doing internal manufacturing, tomorrow you will maybe outsource it to a CMO. Today you have full data in your hands, tomorrow it will go to a paper batch record coming from them, right? Does that mean that your business process will change? No, it should not. You have to still run the business. You still have to provide all the documentation that you need to the regulatory authorities, right?”
But this now shields you from that complexity. Having that particular knowledge graph, the knowledge graph itself is independent of the systems. It is dependent upon your process. It is dependent upon your unit operations. It is dependent upon what nature of data that you create, in what order you are doing them. As long as you’re doing them in a particular order, it will reflect it.Whether you’re doing it in your own site or you’re doing it at a CMO, it really doesn’t matter. If you compare this with the human brain, right? The human brain has about 86 billion neurones and about 100 trillion synapses. So neurones, you can almost conceptualise them as your data and synapses is the relationships, right? In terms of bringing out the intelligence from your data, that is hidden in the relationships.
In the traditional way of dealing with data, you usually store the data in its own form, maybe in a data lake, all of it together. That’s all your neurones, right? But in order to solve any problem, you have to first create the linkages between them. How do they link with each other? Because that’s where the story, that’s where the insight can come from. Our knowledge graph basically builds those insights. It builds those synapses between the neurones in a highly automated way. That’s our AI engine, right? So when we go and connect to any of the data sources, it basically pulls it, it starts connecting it, it starts creating the synapses or the connections between the different data points, and that basically brings meaning to the data.
/ ANY QUESTIONS? WE'D LOVE TO HELP
Frequently Asked Questions
Mareana has helped numerous firms in the Pharmaceutical, Chemical, Medical Device, and Industrial Manufacturing industries.
A manufacturing knowledge graph is a process-driven framework that connects data, systems, and manufacturing operations into a unified structure. Instead of relying on a predefined data model, it reflects the actual manufacturing workflow, making data easier to organize, analyze, and trace.
A knowledge graph improves manufacturing data management by connecting information from multiple sources such as MES, LIMS, ERP, QMS, PAT equipment, and paper records. This creates a single source of truth that enhances visibility, traceability, and decision-making across operations.
Yes. A knowledge graph can integrate data from MES, LIMS, ERP, QMS, PAT systems, and even paper-based records. This enables manufacturers to unify structured and unstructured data without disrupting existing processes.
AI generates insights by automatically identifying relationships between manufacturing data points. By connecting process data, quality records, equipment information, and operational workflows, AI can uncover patterns, improve traceability, and support faster decision-making.
A knowledge graph maintains a consistent process view regardless of whether manufacturing is performed internally or by a Contract Manufacturing Organization (CMO). This helps preserve traceability, compliance, and operational visibility across different manufacturing environments.
A data lake stores large amounts of raw data, while a knowledge graph connects that data through meaningful relationships. By adding context and structure, a knowledge graph makes it easier to discover insights, track dependencies, and understand how manufacturing processes are connected.
Contact Us
Would you like to learn more about our pharmaceutical manufacturing solutions? Fill out the form below and one of our experts will reach out to you promptly.