Devlane Blog
Data Engineering

Data Engineering Trends in 2023

A few years ago, having a data-driven decision-making process was reserved for multinational companies. With the adoption of cloud computing and the ever-increasing democratization of technology, companies of all sizes can generate and analyze vast amounts of data.

Nicolás Mosconi
February 29, 2024

Many companies with privately hosted servers are moving their data, analytics, and use cases to the cloud; its adoption will continue to grow in 2023, with some accelerators such as implementing 5G networks and new methodological approaches like DataOps.

This article will cover some of the most important trends for the upcoming year.

DataOps and MLOps

As it happened with DevOps (Development + Operations), QAops (Quality Assurance + Operations), 2023 will increase the adoption of DataOps and MLOps, for Data Engineering and Machine Learning teams.

The DataOps approach implements the same benefits of DevOps and QAOps for Data professionals. It helps build and maintain streamlined, agile data analytics pipelines iteratively and incrementally. 

Data in businesses is ingested, prepared, and orchestrated in different ways. This creates an increasing demand for automation and pipeline integration tools that help data teams visualize their workflows in unified tools. 

This year, businesses will start implementing DataOps processes, which, combined with the already adopted Agile methodologies, will help companies reach next-level data analytics and decision making.

The DataOps approach will enable organizations to implement automated data pipelines in their private, multi-cloud, or hybrid environments.

The main objective of DataOps and MLOps is to accelerate the development and maintenance cycle of analytics and data models.

Traditionally reserved for big companies, mid-sized companies will start implementing Data and MLOps processes in 2023. Even start-ups with their typical urgency for reliable data will begin implementing some aspects of these new data-driven engineering approaches.

Augmented Analytics

Augmented Analytics (or simply AA) refers to data analytics tools and processes that rely on Artificial Intelligence (AI), Machine Learning (ML), and Natural Language Processing (NLP) techniques. Traditionally handled by a data engineer or a data scientist, AA systems will deliver real-time automated insights.

The results of AA promise to be more accurate, leading to better decisions. With this approach, Data experts can now focus on exploring and generating in-depth reports and predictions. 

Augmented analytics will more than likely experience massive growth in 2023. Augmented analytics also enables non-technical people to rip the benefits of data analytics technology.

AA helps technical and non-technical users understand data through visual AI-assisted analytics. In particular, this technology is a perfect match for online retail platforms that usually generate massive amounts of data that needs processing and analysis.

Multi-Cloud and Hybrid cloud environments

According to Gartner, public cloud services investment will grow more than 20% this year. Companies worldwide move towards hybrid, multi-cloud, and edge environments. This new cloud hosting paradigm enabled innovative distributed data processing architectures.

Organizations moving from private infrastructure to public, hybrid, and multi-cloud solutions will experience a significant increase in agility. Behind this phenomenon lies a connection to a rise in the production of unstructured data.

Using traditional ETL-based batch processing is a thing of the past. Companies now need agile distributed infrastructures that enable them to produce and analyze massive amounts of unstructured business data for improving their decision-making process. 

Organizations are changing the way they look at Big Data and cloud computing.

ML and AI-assisted infrastructures offer better scalability, security, and governance at a lower operational cost.

Hybrid cloud approaches will see an increase in adoption in 2023. By adopting a hybrid cloud, an organization can keep its sensitive data private while combining it with more cost-effective public cloud solutions.

This is especially true for many SMEs. Hybrid clouds for data processing combine cost and security in a balanced and agile way.

Big Data & the Internet of Things (IoT)

The term Internet of Things (IoT) refers to a massive network of small hardware devices that constantly stream data through the web. By integrating IoT with ML and Data analytics, companies can improve their platforms' scalability, flexibility to change, accuracy, and response times.

Small and Medium Enterprises are increasingly adding IoT solutions to their value propositions.

Current online platforms already generate massive amounts of data. With IoT, the generated data is several orders of magnitude higher.

This creates the need for agile collecting, tagging, formatting, and analysis of such vast volumes of information. 

In 2022, data engineers worldwide will face the challenge of processing IoT-generated data and producing valuable insights for decision-making processes. 

Natural Language Processing for Conversational Analytics

Natural Language Processing (NLP) was born as a subset of Artificial Intelligence (AI) techniques. Currently, companies use NLP to study and analyze business data. Many analysts believe NLP tools will increase adoption in 2023 through "Conversational Analytics".

With the use of Natural Language Processing, businesses can leverage the benefits of accessing better quality information, thus generating more relevant insights.

Another aspect of NLP is that it enables what is called Sentiment Analysis. Companies use this type of analysis to better grasp their customers' feelings and thoughts towards their products and services (and their competitors).

Businesses that better understand their customers will naturally offer a better customer experience and a higher level of satisfaction.

Natural Language Processing can feed AI decision-making systems in a profitable way that enables AI for a higher level of understanding. AI systems that provide NLP user interfaces are the game-changer for the following years.

Data fabrics for unstructured data analytics

The traditional data handling approach usually refers to collecting and storing information strictly but rarely uses it. New Agile data handling approaches enable businesses to pull relevant information from massive amounts of unstructured data stored in distributed cloud locations.

The need for unstructured data analytics will see a significant increase in 2023. Companies need to quickly scan their raw unstructured data to find meaningful decision-making insights.

Historically, business intelligence was about designing and feeding data to a structured data warehouse solution, which needed data professionals to develop ways to structure raw data. As IoT, AI, and ML adoption exponentially increases year by year, almost all of the worlds' data is becoming unstructured. 

Data professionals will need to improve their skills and toolsets to handle unstructured data. Companies need to gain insight from data that has no specific structure or predefined schema and that comes in many different supports from video recordings to IoT sensor data to chats and emails.

These many diverse and unstructured sources of data call for storage-agnostic solutions of what are called "data fabrics". 

A data fabric is an architectural pattern designed to move, replicate, access, and visualize data across different cloud storages and resources, enabling near real-time analytics.

IT operations teams involved in DataOps approaches will help choose the proper data fabric architectures for each company, considering sensitive aspects such as Security and Compliance. 


This year will see a massive increase in some already-established approaches like ML and novel methodologies like DataOps. The increasing adoption of multi-cloud and hybrid environments calls for new data fabric architectures that enable businesses to have real-time insight into their data, regardless of storage locations.

Companies need to improve their decision-making processes by adopting better AI-backed analytic tools. Data engineering will be in the spotlight in 2023; the demand for data engineering, data analysts, and AI specialists will increase.

In this article, we have covered some of the most exciting trends in data engineering for 2023:

  • DataOps and MLOps. These methodologies convey the same benefits DevOps brought to the software development field, to Data Engineering and Machine Learning teams. Mid-sized companies will start to implement DataOps this year.

  • Augmented Analytics. With the considerable amount of unstructured data the world generates today, there is a pressing need for better analytic tools. Augmented Analytics (AA) consists of AI, ML, and NLP-improved analytical tools that help technical and non-technical users gain meaningful insight into vast amounts of data.

  • Multi-Cloud and Hybrid environments. Organizations will continue to migrate from private clouds to public or hybrid approaches. Data engineering teams will have to develop storage and transport strategies that both leverage the cost savings of public clouds and have the privacy and compliance benefits of private infrastructures. Businesses worldwide need new Agile-managed distributed infrastructures to handle massive data volumes in a distributed way.

  • Big Data & the Internet of Things (IoT). The amount of data generated worldwide has exponentially increased by the growth of IoT solutions. Millions of devices connect every day to the network to stream massive volumes of unstructured data such as sensors, actuators, and others. There is a need for unified dashboards and tools to analyze this data.

  • Natural Language Processing for Conversational Analytics. Using natural language interfaces, both end-users and Data Engineers will improve how they interact with platforms. Data retrieval and analysis through "conversational analytics" will increase in 2023. Sentiment Analysis, an already established practice, will continue to grow alongside these new NLP-enabled tools.

  • Unstructured data analytics. The field of Data Engineering has traditionally dealt with schemas and structured data. Given the enormous amount of data generated worldwide, data teams need to shift their focus to analyzing data that comes in many shapes and sizes. There will be an increased need to gain insights from data with no specific schemas, from streaming videos to medical information. The diversity of data sources will boost the need for new Data Fabric architectures that enable data analysis in a storage-agnostic way.

Does your company need help in gaining insights from structured and unstructured data? Drop us a line!