Information Supply Chain: the foundation for your data future

Data, data, data.

It’s hard to recall a company who does not have "being data-driven", "utilizing their data better", or a similar goal on their strategy.

Even without mentioning the data, most business strategies call for capabilities that are grounded in it:

• Process optimization (using AI and automation)
• Sustainability (relying on advanced analytics and simulations to assess the current and desired state)
• Better services (based on digital products and customer interaction automation)
• ?Cost-cutting (driven by AI models to define the optimal input/output/cost triangle)

Spoken or unspoken, data is everywhere.

And everywhere it is: stored and duplicated in many systems, hiding in custom-built Excels, carried in the heads of the experts... Often, only the tip of the data iceberg is available through a central Data Platform - and most of the time, it is only available to those who know how to find it.

In this edition, we will cover the emerging concept that will accelerate and de-risk the data-driven business strategies while opening up new exciting business opportunities at the same time: Information Supply Chain.

What is Information Supply Chain

Information Supply Chain describes the process from stating the need for certain information to support a business goal to the delivery of that information in digital format (as data) and contributing the new insights into the pool of information where others can find it. As any Supply Chain, Information Supply Chain relies on multiple processes supporting its core flow:

• Information and Data Compliance: ensuring the Supply Chain does not violate any risk, compliance, legal, or ethical limitation of data access and use.
• Data Quality: ensuring the data is delivered from the source whose quality spec is fit for purpose the data will be used for.
• Knowledge Crowdsourcing: engaging the whole organization into ensuring the Information Supply Chain works at its best.
• Integration Prioritization: ensuring the information that is requested more often gets priority on the backlog of the teams working on the Data Lake (or similar central source for non-operational data needs)

Information Supply Chain Step By Step

Finding Information

Let's say, as a business expert, I have an idea to simplify decision-making on my process based on data.

Or the regulator needs additional data points reported upon.

Or, as a Data Scientist, I suspect there are extra factors influencing the outcome of my models.

What kind of information is available to me? What else is there that could be interesting? Are there things outside of my familiar world that could be of use for my goal?

Here, Data Catalogs are useless. As a user, I am exploring information without any regard for systems just yet. I am not looking for datasets - not at this point, anyway. I don't know what's out there and how to look for it. My perspective and language might differ from these used by most of the rest of the organization. The way I talk might lead to misunderstandings once I cross the safe boundaries of my own business domain.

Information Catalog is what I am looking for. Something that can give me answers in context, supporting me in understanding the others' viewpoints and guarding me against mis-interpreting the information made available by others.

Finding information is not about locating a dataset.

It is about understanding what's there, what it means, and how it relates to the rest of the known world. Eventually it is also about understanding how other business domains look at it.

This step of the Information Supply Chain builds a story that I need to know to support my goal:

• "I need information on alternative parts that can be used in manufacturing carriage components, and their suppliers and prices."
• "I need information about how many of my contractors are using electric vehicles to travel to the office."
• "I need information on the exact composition of raw materials and the manufacturing settings at the time these materials were used."

Locating Data

Sometimes just venturing outside of my own known universe and discovering what else is out there is enough.

But most of the time, I would want to get the data points that represent the story I composed.

The whole story could live in a single dataset, or be scattered around the different datasets of my organization. I need to find the best source for it. Multiple factors will come into play here. For example:

• completeness of the story. The more complete the story, the easier the integration.
• priority of source. Ideally, there would be a dedicated platform where data for non-operational purposes would be taken. Many organizations have Data Lakes just for this purpose. However, not all the data ends up in the Lake, and in some cases, the Data Lake has stringent quality requirements that discard data that could be perfectly fit for some processes. Identifying the best source in this case becomes a matter of looking at origin, OLAP-fit, and criticality of the source.
• update frequency and age of data. Depending on the use case, latency of the data can become an issue, and some use cases require the latest data.

Locating data is about finding the source(s) for the story the user needs to get delivered as data.

Delivering data

The next step is to provide the requested data to the user. This can be done by automatically creating a required dataset (usually only possible if all the data points are available in Data Lake), or by creating a data order towards the teams tasked with data delivery.

Here, just "give me the data!" is not enough.

It is important that people or systems tasked with delivery understand:

• whether to deliver the data: a compliance process might be required for requests that have to do with confidential or personal data, or any time ethical data use concerns might arise
• how to protect the data: what are the standard components and processes to use for data anonymization, obfuscation etc
• where exactly to take the data: a field-level integration spec, if available, will save a lot of discovery and analysis time
• who to ask about the data: shall any questions arise, it should be easy to locate the people in charge of the data
• how to reconcile the data; for composite data sets, a way to put its sources together will significantly speed up the process. Here, we are talking about the interoperability layer.

Tracking the data movements

Once the data is delivered, the information it represents will be available in an additional location. It is essential, both for compliance and for re-use purposes, to ensure there is a trace of it.

Next time someone is looking for the same story, they should be able to see our data destination as another place where it can be found.

Making the new information reusable

So, we have created a report to support our decision-making.

Or to answer the questions from the regulator

Or we have embedded additional features into our AI model.

It's time to tell the rest of the organization that this new information exists.

The next user exploring the information universe should be able to find these new insights and use them to compose their own story.

Information Supply Chain vs…

Data Mesh

Information Supply Chain can be implemented on top of Data Mesh.

It will turn your Data Mesh into Knowledge Mesh, where available insights are not limited to the data exposed to other domains, and does not necessarily need to be part of any Data Product.

In addition, the data - including that exposed in Data Products - will be accessible to the user in their own lingo and perspective, avoiding uncertainty and semantic data quality issues.

Data Catalog

Information Supply Chain is about much more than simply cataloging data sources. It supports the business user to find information - even if it is not available in digital format. It also enables the user to compose a story rather than simply collecting datapoints.

Properly implemented Information Supply Chain should support the crowdsourcing of knowledge by enabling every user to see the information universe from their own viewpoint, and therefore to contribute to it with confidence and without needing specialistic training.

Enterprise Data Lake

The Information Supply Chain paradigm does not assume the data is available in a centrally curated location. Instead, it should track the information requests that go through it in order to prioritize the efforts of the central Data team bringing new data into the Lake. In the meantime, it should support the discovery and delivery of data from the operational landscape.

Creating Information Supply Chain

Information Supply Chain done right can accelerate the delivery of datacentric business strategy manyfold.

• It removes the repetitive effort of locating, analyzing, assessing, controlling, and sourcing the data.
• It reduces risks by preventing semantic data quality issues and embedding compliance into the very fabric of data delivery.
• It prevents the over-duplication of data - also called "databesity", therefore saving storage, integration, and information lifecycle costs.

What's most important, it enables everyone in the organization - whatever their role and degree of data proficiency - to use the whole body of knowledge the organization assembled in different domains over the years to automate and innovate.

Traditionally, Information Supply Chain would take a massive effort requiring multiple functions and a whole array of Data Management solutions, as well as user education and change management. This would significantly reduce both the viability and the attractiveness of Information Supply Chain initiatives.

This is changed with Eva by Hakoona.

Eva is a single solution that delivers all aspects of the Information Supply Chain while creating a ready-to-use foundation for your Digital Twin initiatives.

This means the same effort will deliver three benefits:

• end-to-end data foundation, including contextual Business Glossary,? Information Catalog, Data Catalog, 360 degree Metadata Management solution, Master Data Management platform, API generator, and Interoperability Layer generator
• Information Supply Chain, with all its processes and workflows
• ready data backend for business solutions like Digital Twin, Operational Compliance, AI solutions, or Process Optimization initiatives.

Interested in Eva?

It is not too late join our co-creation program to build some of its features with us, and get a whole package of discounts and benefits reserved to Hakoona Diamond Customers.?

Contact us to learn more!