3 culinary techniques for enterprise data chefs
When great chefs and first-tier enterprise architects are at the top of their game, they can create works of art. These experts are the people to watch and learn from as they hone their craft and develop into bona fide artists.
Both groups have a lot in common. Master chefs and data chefs understand a wide array of ingredients and choose from the freshest available. They minimize waste. They know how to follow a recipe—and more importantly, can adjust on-the-fly when needed. Both are also incredibly creative—understanding how to use ingredients in multiple ways.
Chocolate in chili? Yes! AI mixed with OLTP and OLAP? Yes! And finally, they stay current with the latest trends, picking and choosing those that align with their sensibilities.
Alongside artistry, both master chefs and data chefs are scientists who are always under pressure to be innovative and add techniques and recipes that produce high quality, consistent results. If you’re still not sure what a master chef’s perfectly cooked steak has to do with data architect’s big data warehouse, here are three recipes that will give you food for thought.
Orchestrate Data for Repeatable, Consistent Outcomes
What makes a recipe a real winner is repeatability. Can you recreate it again? Think about a perfectly cooked steak. Chefs use the sous vide technique to cook and serve a steak perfectly every time. They place the cut of meat in vacuum seal bag, bring the steak to 129 degrees, then sear it, and serve it up with no guesswork around whether it’s done. The steak is always cooked to the exact degree using the sous vide technique, and the chefs can prepare a perfectly cooked steak again and again.
Similarly, data architects need a real-time, in-memory relational database management system and tools that serve up data across the enterprise the same way every time. While businesses have loads of data, they are still hungry for quality data to fuel innovation and business growth. Too often data is unavailable in siloed environments or lacks integrity and many enterprises struggle with their data being from a variety of locations. Bad data, just like bad ingredients will ruin a project’s outcome.
When data is orchestrated across the enterprise, the enterprise data team can help each business unit fulfill their unique goals. Data is stored in a shared environment and can be modeled and governed. It doesn’t need to be wrangled, and the data can be manipulated or coordinated for any internal team that wants to monetize their data.
To orchestrate data and ensure consistent outcomes, follow these data chef directions:
- Get a consolidated view of all data from all data sources, covering business processes and applications.
- Improve data quality through cleansing, and get a graphical view of data correlations and linage across the enterprise.
- Employ distributed data pipeline processing and refinement using a variety of computation techniques such as OLAP, graph, time series and machine learning.
- Orchestrate data end-to-end, process data where it is located, and avoid expensive data movement.
- Maintain security policy dynamically in one place, and help ensure that policy measures are in place to meet regulatory and corporate requirements.
Maintaining Control in a Big Data Warehouse
Recipes and cooking techniques come in and go out of vogue just like IT trends. Think about how Julia Child and client-server technology has made way for Smitten Kitchen and the cloud.
A major new trend in data strategies is around data. Previously, the goal was to have all data collected in one platform. The new strategy is to have a data management solution that can manage all data types from disparate sources and adapt to an evolving IT infrastructure.
Pro data chefs need tools in their Big Data warehouse that give them the control to support analytics and easily handle data ingestion, onboarding, integration and quality requirements across all forms of data sources and types. Basically, what’s needed is an agent that works like yeast.
Yeast is a tricky ingredient, as it has many forms. Left alone, yeast is inactive, but add warm water and it comes alive, releasing gas that makes dough rise. Elite bakers know how to work with yeast to turn a lump of flour into amazing breads and pastries.
What if you had an agent that had a similar effect on the huge volumes of data within your business? Most organizations only use 20 percent of their data. The rest of it is dormant, or a lump of uninteresting dough. In the same vein, data’s value has an expiration date, as does yeast. Businesses that can quickly begin using incoming data will have many more innovative use cases than those tied to less flexible data platforms.
A standard recipe for creating a high value, big data warehouses is:
- Appoint big data warehouse lead.
- Inventory different systems and data types.
- Determine operational business and IT requirements.
- Identify data governance and analytical requirements.
- Define enterprise architecture deployment model.
- Implement data lifecycle and multi-temperature management.
Fusion: The Best of Two Worlds
Fusion is trend that is fun for chefs and diners. Some of the hardest reservations to get these days are at restaurants mixing up Japanese traditions with California farm-to-table sensibilities or French standards with a middle Eastern flair. Similarly, businesses shouldn’t get left out of the innovations being introduced with the combination of geoinformation data with business data.
A southern California hospital has fused geospatial data with business data to help prospective customers understand primary and specialist care coverage. With spatial data in their kitchen, the hospital is converting provider data, hospital data, patient data, geoinformation, and direct-to-employee data into a dashboard of graphs, maps, and charts for multiple business stakeholders.
In a unique turn, personnel from throughout the hospital are using the data for their specific needs. The business development team is filling the sales pipeline, the health management team is identifying gaps and opportunities, and the leadership team is improving patient experience and outcomes.
The directions for serving up this fusion of data flavors is:
- From ArcGIS, ESRI or other geospatial platform connect to in-memory, real-time database via ODBC and create and enterprise GeoDB.
- Copy or publish GIS data into GeoDB created in database platform.
- Create a view in database and join the business data and the GIS data.
- Build visualizations of the data.
- Apply additional spatial functions and optimization to your taste.
- Overly demographic data from ESRI ArcGIS online.
Waste Not, Want Not
One disturbing food fact is that roughly 30 percent of globally produced food is lost or wasted annually. That figure is disturbingly close to the stats around how much data is lost or wasted within an organization. In another similarity, chefs and data architects share the same goals to find a home for all their food and data. Both groups strive for 0-percent waste because they know that throwing away good food and good data serves no one.
Here’s to creating more no-fuss recipes with data as the central ingredient for feeding a hungry business.
(If you are interested in more data chef recipes, go to sap.com/datachef and follow #SAPDataChef.)