Data Buzzwords You Need To Know in 2023 — Part II | by Rashi Desai | Feb, 2023

Data terms you’re likely to come across this year & what they mean

Back in early Jan, I wrote about 13 Data Buzzwords You Need To Know In 2023 with TDS Editors. I resisted typing a 15-minute long read then to include 23 buzzwords for 2023 and yet, I knew I had to write about those 10 buzzwords soon! Through comments on that blog, Twitter DMs, and LinkedIn messages, I heard a few of those 10 words that I should have covered and a few new ones. With that, here is Part 2 to the data buzzwords you should know about in 2023 to catch up on the data vocab!

In 2023, companies will either live or die by their data strategies.

With the increase in the number of data sources, acquisition channels, and insights consumed by a business, an organization must be equipped to scale and handle the changing demands by adding/removing data resources.

Data scalability is the ability of data, databases, or any data systems to increase or decrease the processing, without a significant impact on performance and cost in response to changes in processing demands. Scalability is quintessential today as it allows a business to grow and generate dollars without being held back by its data structure or lack of capacity and resources.

One of the challenges that organizations consuming data face today is the decentralized nature of data management and governance which has led people to talk about data mesh in 2023.

Data mesh takes the approach to build a decentralized data architecture for the decentralized data with organizations. It essentially stitches together data held across multiple data silos (cloud, on-premise data, application database, BI dashboards, analytics applications, etc.

There is often a comparison between data mesh and data fabric — the better option for organizations and let me tell you, they are very different. While a data fabric governs and manages multiple data sources from a single, centralized system, a data mesh does the complete opposite. The end goal remains the same — to connect distributed data across different locations and organizations.

Data lakehouse has been an emerging concept for the past few years and 2023 for me is the year when lakehouses flex their functionality and capabilities (peak of their existence).

I’m sure everyone reading this blog must have heard or read of Databricks atleast once. Databricks is a data lakehouse and what a data lakehouse does is combine the operational excellence, reliability, security, compliance, performance efficiency, cost optimization, and scale of organizational data to enable business intelligence (BI) and machine learning (ML) on all that good company data.

I don’t know about you but I had a question — how are data lakes different from data warehouses?

Both are widely accepted and used for storing big data, but what I understand the difference is — a data lake is a vast pool of raw data with no defined purpose when the data enters whereas, a data warehouse is a structured repository that hosts filtered, processed data for a defined purpose.

Data is the center of any conversation happening in the tech world today and if we talk about how much data is created every day, the current estimate in 2023 stands at 1.145 trillion MB per day.

We collect, process, and store data for “regular business activities” and once the project is done with, generally fail to use that data for other purposes. We define dark data as data acquired through various sources but not used to analyze or derive insights for decision-making.

Storing and securing data incurs more expense to the business and sometimes, a greater risk to the organization than value. Some examples of dark data for an organization could be log files, account information, financial archives, data of former employees, etc. that are no longer useful and are sitting there consuming space.

I heard the word Web3 on a podcast in mid-2022 where a stand-up comedian talked about his interests other than comedy and he talked about Web3, the opportunities that exist with Web3, and the finances involved in its emergence.

I was intrigued by what that meant and googled to read that Web3 is the third generation of the World Wide Web which introduces a decentralized internet backed by blockchain technologies, and token-based economics. We started seeing a Web3-friendly approach in 2022 and in 2023, more and more countries are developing and passing legislation to position themselves as “Web3-friendly” zones.

Web3 is here to completely change how data is stored and used. Instead of data stored and controlled by one central entity in a giant server, data will now be owned by millions of user devices linked together by one network (decentralization) — improving personal data ownership. Web3 employs a machine-based understanding of data to develop more intelligent and connected web experiences for users.

2023 is the year for Data Analysts. One of the pain points for any data analyst is maintaining consistency across all reporting efforts. So, if we want 2023 to be a good year for the analysts, we pivot to synchronization.

Data synchronization is precisely what it sounds like — making sure your data is consistent, identical, and accurate across all applications, regardless of their location (basically, every data analyst’s nightmare).

I read an example in a blog sometime to better understand data synchronization and here’s my hot take to understand the term — say, we added a new album with five songs to one of the servers of a music streaming platform server, so, with data synchronization, all other music streaming platform servers (under an agreement) get identical sets of the album. Data synchronization plays a prime role in database management. Data synchronization also brings data governance and secured access controls.

The worldwide cloud market is expected to grow by 42% in 2023 and I anticipate a mass movement across organizations and industries to optimize cloud costs and spending this year.

Organizations are moving to the cloud, debating staying on-premise or going with a hybrid model, and amongst all those good conversations, a distributed cloud emerges as an option where multiple clouds are used by an organization to meet the performance expectations, compliance needs, and support computing requirements while being centrally managed from a public cloud provider.

I confused distributed cloud with the hybrid cloud when I first learned about it and wanted to put together recommendations for my workplace here’s how they differ — hybrid cloud consists of multiple public cloud services with on-premises infrastructure whereas distributed cloud is managed by one central public cloud provider.

Automation of everyday jobs has been a priority for a lot of organizations in 2023 (so is for my team) and hyper-automation targets to automate everything in an organization that can be.

Hyperautomation aims to streamline data processes, artificial intelligence (AI) jobs, robotic process automation (RPA), and other technologies to run autonomously without any human intervention. Hyperautomation starts by identifying what projects to automate, choosing the appropriate automation tools, driving swiftness through multiple automated processes, and extending their capabilities using various flavors of data.

Human-Centered AI as a word has been floating around for a while now. Everything we do and create today does not operate in a void. It will involve a human at some point of interaction. I’ve always been curious to see what would a practical implementation of Human-Centered Data Science look like and there are a few projects surfacing in 2023 where human interactions with the product lead the business decisions.

I like to define Human-Centered Data Science as a concept to understand how people interact and make sense of social situations, enabling humans to explore and gain insights and design data models with the end-user in mind (not just the business).

Let’s look at it this way — data is nothing but the digital traces of human interactions. A human-centered approach can enhance the choices data scientists make every day, by making the process more transparent, asking questions, and considering the social context of the data.

Decision Intelligence is the new-age, smart decision-processing method.

Observe > Investigate > Model > Contextualize > Execute

I first came to know about Decision Intelligence (DI) in 2021 during my internship with PepsiCo where one of the projects I came across was to use decision intelligence and forecast the sales of Pepsi food products across demographics before and after the Super Bowl. DI offers a clean and neat structure for data-driven decision-making.

You use your descriptive, diagnostic, and prescriptive analytics and analyze data each with its own set of characteristics being the core force. With decision intelligence, you create 3 different types of models: human-based decisions, machine-based, and hybrid decisions. The process involves a heavy reduction algorithm that combines data science with social science, decision theory, and managerial sciences.

I’ve read an emerging number of job descriptions that cater to psychology majors with an interest in analytics to work on decision intelligence by the day.

That’s it from my end on this long blog post. Thank you for reading! I hope you found it an interesting read. Let me know in the comments about your experience with storytelling, your journey in data, and what are you looking for in 2023!

If you enjoy reading stories like these, consider signing up to become a Medium member from this link.

Happy Data Tenting!

Rashi is a data wiz from Chicago who loves to visualize data and create insightful stories to communicate insights. She’s a full-time healthcare data analyst and blogs about data on weekends with a cup of hot chocolate…

Data terms you’re likely to come across this year & what they mean

Back in early Jan, I wrote about 13 Data Buzzwords You Need To Know In 2023 with TDS Editors. I resisted typing a 15-minute long read then to include 23 buzzwords for 2023 and yet, I knew I had to write about those 10 buzzwords soon! Through comments on that blog, Twitter DMs, and LinkedIn messages, I heard a few of those 10 words that I should have covered and a few new ones. With that, here is Part 2 to the data buzzwords you should know about in 2023 to catch up on the data vocab!

In 2023, companies will either live or die by their data strategies.

With the increase in the number of data sources, acquisition channels, and insights consumed by a business, an organization must be equipped to scale and handle the changing demands by adding/removing data resources.

Data scalability is the ability of data, databases, or any data systems to increase or decrease the processing, without a significant impact on performance and cost in response to changes in processing demands. Scalability is quintessential today as it allows a business to grow and generate dollars without being held back by its data structure or lack of capacity and resources.

One of the challenges that organizations consuming data face today is the decentralized nature of data management and governance which has led people to talk about data mesh in 2023.

Data mesh takes the approach to build a decentralized data architecture for the decentralized data with organizations. It essentially stitches together data held across multiple data silos (cloud, on-premise data, application database, BI dashboards, analytics applications, etc.

There is often a comparison between data mesh and data fabric — the better option for organizations and let me tell you, they are very different. While a data fabric governs and manages multiple data sources from a single, centralized system, a data mesh does the complete opposite. The end goal remains the same — to connect distributed data across different locations and organizations.

Data lakehouse has been an emerging concept for the past few years and 2023 for me is the year when lakehouses flex their functionality and capabilities (peak of their existence).

I’m sure everyone reading this blog must have heard or read of Databricks atleast once. Databricks is a data lakehouse and what a data lakehouse does is combine the operational excellence, reliability, security, compliance, performance efficiency, cost optimization, and scale of organizational data to enable business intelligence (BI) and machine learning (ML) on all that good company data.

I don’t know about you but I had a question — how are data lakes different from data warehouses?

Both are widely accepted and used for storing big data, but what I understand the difference is — a data lake is a vast pool of raw data with no defined purpose when the data enters whereas, a data warehouse is a structured repository that hosts filtered, processed data for a defined purpose.

Data is the center of any conversation happening in the tech world today and if we talk about how much data is created every day, the current estimate in 2023 stands at 1.145 trillion MB per day.

We collect, process, and store data for “regular business activities” and once the project is done with, generally fail to use that data for other purposes. We define dark data as data acquired through various sources but not used to analyze or derive insights for decision-making.

Storing and securing data incurs more expense to the business and sometimes, a greater risk to the organization than value. Some examples of dark data for an organization could be log files, account information, financial archives, data of former employees, etc. that are no longer useful and are sitting there consuming space.

I heard the word Web3 on a podcast in mid-2022 where a stand-up comedian talked about his interests other than comedy and he talked about Web3, the opportunities that exist with Web3, and the finances involved in its emergence.

I was intrigued by what that meant and googled to read that Web3 is the third generation of the World Wide Web which introduces a decentralized internet backed by blockchain technologies, and token-based economics. We started seeing a Web3-friendly approach in 2022 and in 2023, more and more countries are developing and passing legislation to position themselves as “Web3-friendly” zones.

Web3 is here to completely change how data is stored and used. Instead of data stored and controlled by one central entity in a giant server, data will now be owned by millions of user devices linked together by one network (decentralization) — improving personal data ownership. Web3 employs a machine-based understanding of data to develop more intelligent and connected web experiences for users.

2023 is the year for Data Analysts. One of the pain points for any data analyst is maintaining consistency across all reporting efforts. So, if we want 2023 to be a good year for the analysts, we pivot to synchronization.

Data synchronization is precisely what it sounds like — making sure your data is consistent, identical, and accurate across all applications, regardless of their location (basically, every data analyst’s nightmare).

I read an example in a blog sometime to better understand data synchronization and here’s my hot take to understand the term — say, we added a new album with five songs to one of the servers of a music streaming platform server, so, with data synchronization, all other music streaming platform servers (under an agreement) get identical sets of the album. Data synchronization plays a prime role in database management. Data synchronization also brings data governance and secured access controls.

The worldwide cloud market is expected to grow by 42% in 2023 and I anticipate a mass movement across organizations and industries to optimize cloud costs and spending this year.

Organizations are moving to the cloud, debating staying on-premise or going with a hybrid model, and amongst all those good conversations, a distributed cloud emerges as an option where multiple clouds are used by an organization to meet the performance expectations, compliance needs, and support computing requirements while being centrally managed from a public cloud provider.

I confused distributed cloud with the hybrid cloud when I first learned about it and wanted to put together recommendations for my workplace here’s how they differ — hybrid cloud consists of multiple public cloud services with on-premises infrastructure whereas distributed cloud is managed by one central public cloud provider.

Automation of everyday jobs has been a priority for a lot of organizations in 2023 (so is for my team) and hyper-automation targets to automate everything in an organization that can be.

Hyperautomation aims to streamline data processes, artificial intelligence (AI) jobs, robotic process automation (RPA), and other technologies to run autonomously without any human intervention. Hyperautomation starts by identifying what projects to automate, choosing the appropriate automation tools, driving swiftness through multiple automated processes, and extending their capabilities using various flavors of data.

Human-Centered AI as a word has been floating around for a while now. Everything we do and create today does not operate in a void. It will involve a human at some point of interaction. I’ve always been curious to see what would a practical implementation of Human-Centered Data Science look like and there are a few projects surfacing in 2023 where human interactions with the product lead the business decisions.

I like to define Human-Centered Data Science as a concept to understand how people interact and make sense of social situations, enabling humans to explore and gain insights and design data models with the end-user in mind (not just the business).

Let’s look at it this way — data is nothing but the digital traces of human interactions. A human-centered approach can enhance the choices data scientists make every day, by making the process more transparent, asking questions, and considering the social context of the data.

Decision Intelligence is the new-age, smart decision-processing method.

Observe > Investigate > Model > Contextualize > Execute

I first came to know about Decision Intelligence (DI) in 2021 during my internship with PepsiCo where one of the projects I came across was to use decision intelligence and forecast the sales of Pepsi food products across demographics before and after the Super Bowl. DI offers a clean and neat structure for data-driven decision-making.

You use your descriptive, diagnostic, and prescriptive analytics and analyze data each with its own set of characteristics being the core force. With decision intelligence, you create 3 different types of models: human-based decisions, machine-based, and hybrid decisions. The process involves a heavy reduction algorithm that combines data science with social science, decision theory, and managerial sciences.

I’ve read an emerging number of job descriptions that cater to psychology majors with an interest in analytics to work on decision intelligence by the day.

That’s it from my end on this long blog post. Thank you for reading! I hope you found it an interesting read. Let me know in the comments about your experience with storytelling, your journey in data, and what are you looking for in 2023!

If you enjoy reading stories like these, consider signing up to become a Medium member from this link.

Happy Data Tenting!

Rashi is a data wiz from Chicago who loves to visualize data and create insightful stories to communicate insights. She’s a full-time healthcare data analyst and blogs about data on weekends with a cup of hot chocolate…