What does data transformation enable data analysts to accomplish?

Author Image

Mikołaj Kromka

Principal Software Engineer

15 minutes read

Many organizations collect information to make data-driven decisions. But raw data, freshly collected, is seldom ready for analysis. Data comes in diverse formats, structures, and sizes. Preprocessing this data is essential to make it usable, coherent, and insightful. This preprocessing is known as data transformation, typically conducted through Extract, Transform, Load (ETL) or Extract, Load, Transform (ELT) methods.

Data transformation is crucial in converting raw data into valuable insights. It equips data analysts—and ultimately businesses—to achieve their objectives. Here’s what data transformation accomplishes:

  • Enhances data quality and consistency – Data transformation rectifies inconsistencies, errors, and missing values in raw data, rendering it more reliable and accurate.
     
  • Facilitates interpretation and analysis – Data transformation structures data in a way that makes it accessible and easy to interpret. Analysts then extract meaningful patterns, trends, and relationships.
     
  • Enables compatibility with downstream processes – It tailors the data to meet the unique demands of downstream processes, such as BI reporting, ML, and ad hoc analytics. 
     
  • Supports real-time decision-making – Data transformation optimizes data for real-time processing, enabling organizations to extract insights on the fly and respond swiftly to changing situations. 
     
  • Enhances scalability and performance – Data transformation enhances the performance of data processing pipelines. Whether it’s distributed processing using frameworks like Apache Spark or cloud-based solutions like BigQuery.
     
  • Empowers cross-functional collaboration – It bridges the gap between technical and non-technical stakeholders within an organization. Analysts, data scientists, and business users collaborate effectively with standardized and transformed data.

These benefits significantly advance you and your organization. Yet, several challenges must be navigated to utilize data transformation effectively. We’ll examine various transformation techniques, discuss selecting the appropriate technology, and delve into some methods in detail. 

What allows the transformation of data into information

Data transformations shape vast and complex datasets into clear, actionable information. However, the landscape of transformations is far from one-size-fits-all. It’s a multifaceted realm where various methods and technologies blend.

What_does_data_transformation_enable_data_analysts_to_accomplish_yoast_Graph_1

We’ll embark on a short journey of exploration to unravel the complexities of data transformation. We’ll delve into the three methods enabling transformations and briefly break them down to illustrate their possibilities.

SQL-based data transformations

SQL is a traditional, declarative language for transforming structured data. It is a battle-tested approach in data warehousing with almost 50 years of history. SQL-based data transformations involve using SQL statements to perform operations on data. These operations include filtering, sorting, aggregating, and joining tables. You specify the desired outcome, and the execution engine determines how to execute it.

  • Custom SQL queries in relational databases allow analysts to write custom queries tailored to specific needs. Data analysts filter, join, aggregate, and manipulate data in databases much faster and more efficiently. 
     
  • Generic data transformations with dbt: Data Build Tool (dbt) is a SQL-based transformation framework that facilitates modular, generic, version-controlled, and tested transformations. Analysts then create reusable models, ensuring consistency and maintainability in transformations.
     
  • Cloud Data Warehouses: Cloud-based platforms like Snowflake and BigQuery offer powerful SQL capabilities at scale. They enable parallel processing of large datasets, allowing for complex data transformations in distributed environments.

Programming language-based data transformations

Unlike SQL, a declarative language, imperative languages offer a broader set of tools and functionalities for data transformations. Developers explicitly dictate the calculation process, offering greater flexibility and opportunities for crafting generic and reusable code. 

This distinction empowers developers to abstract away implementation details and focus on expressing high-level objectives in declarative languages while providing more granular control over operations in imperative languages. Among those, you will find Python, Java, and Scala.

  • Python Libraries: Python offers a plethora of libraries for data manipulation and analysis. Let’s have a look at some of the most important ones:
    • Pandas provides versatile data structures for efficient data transformation
    • Dask enables parallel computing for larger datasets
    • PySpark integrates with Apache Spark, enabling distributed data processing
    • Polars and Ray offer high-performance parallel processing options
       
  • Java-based frameworks: Apache Spark and Apache Flink offer distributed data processing capabilities. They allow transformations on vast datasets, supporting real-time and batch-processing scenarios. This makes them ideal for big data applications.
     
  • Scala with Apache Spark and Kafka Streams: Scala provides a concise and expressive syntax for distributed data processing when integrated with Apache Spark. Also, with Kafka Streams developers create robust stream processing pipelines. It allows the most concise, error-prone, and maintainable code. 
     
  • R: is a language specifically designed for data analysis and statistics, and offers packages like dplyr as well as tidyverse. These packages provide an extensive set of functions for data transformation, manipulation, and analysis.

GUI-based tools for data transformations

These are applications with graphical user interfaces that allow users to visually design, configure, and execute data transformation workflows without writing code. They simplify the transformation definition process and make it accessible to users with varying technical expertise.

  • Dataiku: Dataiku is a collaborative data science platform that offers a visual interface for designing data pipelines. It provides drag-and-drop components for transforming data. This makes it accessible for both technical and non-technical users.
     
  • NiFi: Apache NiFi enables the automation of data flows across systems. Its intuitive UI allows users to design, control, and monitor data flows, making it ideal for real-time data transformations and routing.
     
  • KNIME: is an open-source data analytics, reporting, and integration platform with a visual workflow designer. It allows users to create data transformation workflows using a variety of nodes. This makes it easy to manipulate, enrich, and preprocess data without writing code.

Data Transformation: ETL and ELT Techniques

While we’ve discussed various methods like SQL-based data transformations, programming language-based data transformations, and GUI-based tools, it’s important to note that these methodologies often align with different data processing techniques such as ETL and ELT.

  • ETL, which stands for Extract, Transform, Load, involves extracting data from different sources, transforming it into a usable format, and then loading it into a target system.
What_does_data_transformation_enable_data_analysts_to_accomplish_yoast_Graph_2
  • ELT or Extract, Load, Transform performs the extraction of data, loading it into a target system, and then transforming it as needed.
What_does_data_transformation_enable_data_analysts_to_accomplish_yoast_Graph_3

For instance, the process of ETL from operational platforms (OLTP) to analytical platforms (OLAP) illustrates this transition. Online Transactional Processing (OLTP) systems are designed to capture and manage real-time transactional data, whereas Online Analytical Processing (OLAP) systems specialize in organizing and analyzing extensive datasets to facilitate decision-making.

When considering its alignment with transformation methods, we observe the following pattern: SQL-based data transformations might be prevalent in traditional ETL processes, whereas programming language-based data transformations might find relevance in ELT and Big Data platforms, allowing for more agile and scalable data processing.

When considering ETL processes within SQL-based data transformations, organizations often utilize platforms like Snowflake or BigQuery. These platforms offer powerful SQL capabilities and scalable infrastructure, facilitating the extraction, transformation, and loading of data into analytical repositories like OLAP systems.

When discussing programming language-based data transformations, ELT methodologies can seamlessly integrate with Python libraries like Pandas, PySpark, or Scala integrated with Apache Spark. These languages and tools empower developers to perform extraction and loading while enabling powerful transformations, analysis, and machine learning models, as OLAP systems demand.

Data transformations form the backbone of converting raw data into meaningful insights. Whether through traditional ETL or modern ELT methodologies, integrating OLAP systems for analysis, or connecting OLTP systems for real-time data extraction, the choice of approach and technology plays a critical role in shaping the organization’s data strategy. 

The pros and cons of data transformation methods

The right choice of data transformation methods is crucial to success. However, several factors influence the decision of choosing one. Selecting the right method may be based on your team’s specific requirements, technical expertise, complexity of data transformations, and the desired level of customization.

For instance, consider an organization’s familiarity with a particular technology. Opting for a complete change in technology could prevent the acceptance and integration of the new solution. In this case, improving the existing solution with smart engineering techniques seems to be the better choice

But let’s go back on track and look at the different data transformation methods. SQL-based data transformations offer simplicity and speed. Whereas programming language-based approaches provide unparalleled flexibility. GUI-based tools, on the other hand, promise user-friendly interfaces. 

As with any technology, each approach has its pros and cons. To illustrate them coherently, we will look at tables outlining the advantages and disadvantages.

Keep in mind that the following tables present challenges for organisations without increased data engineering knowledge or limited resource capacity. If need be, you can always rely on experts like software engineering partners to avoid pitfalls or increased costs in the long run. They create tailored data solutions efficiently.

SQL-based data transformations

SQL-based transformations are ideal for tasks involving structured data, routine data processing, and scenarios where simplicity and ease of use are prioritized over complex custom logic.

Pros of SQL-based data transformationsCons of SQL-based data transformations
Simplicity and Versatility:
SQL offers a straightforward syntax for querying databases that makes it user-friendly. It allows users to perform data transformations, from basic filtering to complex joins and aggregations.
Limited Complexity:
SQL might struggle with complex transformations or operations that require custom logic. This limits its applicability in scenarios requiring intricate data manipulations.
Optimised Performance:
SQL engines are optimised for query execution, ensuring fast processing times, especially for structured data and standard operations.
Vendor-specific Syntax:
Different database systems often have vendor-specific SQL dialects, which can lead to portability issues when migrating queries across platforms.
Scalability:
SQL queries can scale horizontally, enabling efficient processing of large datasets when used with distributed databases like BigQuery and Snowflake.
Difficult verification of results and debugging: 
Debugging complex queries or creating effective automated tests can be challenging due to SQL’s declarative nature and the absence of typical software engineering tools.

Programming language-based data transformations

Programming language-based data transformations are well-suited for complex tasks, involving custom algorithms, machine learning, and scenarios demanding extensive data manipulation capabilities.

Pros of programming language-based data transformationsCons of programming language-based data transformations
Flexibility
Programming languages like Python and Scala offer immense flexibility. They allow developers to implement custom algorithms, machine-learning models, and complex data manipulations tailored to specific use cases.
Learning Curve
Programming languages require proficiency, making them less accessible for non-technical users. Learning to code and understanding language-specific nuances can be time-consuming.
Community Support
These languages benefit from vast libraries and active communities. Python, for instance, has pandas for data manipulation, scikit-learn for machine learning, and many more, enhancing its capabilities.
Resource Intensive
Custom programming may be resource-intensive, particularly for large-scale data processing, which can increase costs and infrastructure needs.
Extensibility
Programmatic transformations easily incorporate external libraries, tools and custom, self-made generalisations. This enables seamless composition of diverse functionalities into data processing workflows.
Complexity in Standardisation
Due to the freedom in defining transformations, where multiple approaches exist, ensuring diligence and promoting good practices becomes essential for sharing knowledge across the team.

GUI-based tools for data transformations

GUI-based tools are ideal for scenarios where collaboration between technical and non-technical teams is essential, rapid prototyping is required, and the transformations are relatively straightforward and well-suited for the tool’s capabilities.

Pros of GUI-based tools for data transformationsCons of GUI-based tools for data transformations
User-Friendly
GUI-based tools offer intuitive interfaces, allowing non-technical users to design data workflows without writing code. It promotes collaboration between technical and non-technical teams.
Limited Customisation
GUI tools might have limitations in implementing highly customised or specialised transformations, restricting flexibility for advanced use cases.
Visual Representation
Graphical interfaces visually represent data pipelines, making it easier to understand, modify, and optimise workflows, enhancing overall transparency.
Scalability Challenges
Some GUI-based tools may face scalability issues with large datasets or complex transformations, affecting performance in high-volume processing scenarios.
Rapid Prototyping
Users can quickly prototype and iterate transformations without diving into complex coding, enabling faster development cycles.
Long-term maintenance burden
Performing significant alterations, like refactoring or generalisation, to the project via the GUI interface can become time-consuming, error-prone or, in some cases, unfeasible.
 Challenges due to the non-textual nature
Implementing good engineering practices, like conducting code reviews to compare versions, maintaining consistent code style, performing targeted unit tests, or automating various stages of the development cycle, becomes challenging or unattainable due to the absence of text-based content.

Each approach has distinct strengths and weaknesses, offering suitable solutions for various data transformation situations. By understanding these strengths, we can move to choosing the right technology for your project.

Choosing the right technology for data transformations

Choosing the right technology involves carefully considering several crucial factors. Each plays a pivotal role in ensuring the seamless flow of data and the effective data transformation process. In the end, we want to get quality data that improves the decision-making process, instead of patching holes and eliminating errors.

To be able to choose correctly, we need to consider 8 factors. Let’s break them down and tackle each characteristic individually.

Factor 1 – Data characteristics

Understanding your data is crucial for helpful data transformations. The table below highlights key aspects of data: its structure, size, and storage location. By knowing your data, you move closer to accurate and efficient data transformations.

CharacteristicConsiderationCriteria
Data structureIdentify if the data is structured like tables or columns, or unstructured data, such as JSON files or multimedia formats.Identify if the data is structured like tables or columns or unstructured data, such as JSON files or multimedia formats.
Data sizeDetermine the volume of data you need to process and analyse.Opt for scalable solutions like Apache Spark or cloud-based services like AWS EMR for large datasets. In-memory processing libraries like pandas in Python or R can provide fast and efficient data transformations for smaller datasets.
Data storageIdentify where the data is stored, such as cloud storage (AWS S3, Google Cloud Storage), IoT devices, or traditional relational databases.Choose tools compatible with your storage infrastructure. For cloud storage, technologies like Apache Beam and Apache Spark can be used. Real-time streaming frameworks like Apache Kafka can handle high-volume data streams for IoT devices.

Factor 2 – Complexity of data transformation

Understanding complex data transformations requires knowing the details of customization and complexity. This table explains important factors for assessing transformation complexity, from aggregations to machine learning models and custom logic.

CharacteristicConsiderationCriteria
ComplexityEvaluate the complexity of data transformations required, including aggregations, machine learning models, and custom logic.Select tools that support complex data transformations, such as TensorFlow for machine learning tasks and custom processing using Python or Scala in Apache Spark for intricate data manipulations.
Customisation needsDetermine if transformations demand specialised, customised logic tailored to specific use cases.Opt for programming languages like Python, Scala, or Java, that allow custom code implementations. For example, libraries like PySpark and scikit-learn (Python) provide extensive customisation options for machine learning models. For Scala you might want to use Apache Spark or Flink.

Factor 3 – Data processing requirements

Knowing how often data arrives and how quickly it’s needed is crucial, because it leads to timely insights and better decision-making.

CharacteristicConsiderationCriteria
Data FrequencyAnalyse the frequency of data arrival — real-time streaming or batch processing.Choose technologies suitable for the data velocity. For real-time processing, Apache Kafka Streams and Apache Flink are ideal. For batch processing, Apache Spark and Hadoop can efficiently handle large volumes of data. Or maybe, you need to apply a technology for both velocities.
LatencyDetermine the acceptable delay in data processing.Opt for real-time processing frameworks like Kafka Streams or Apache Flink, which offer minimal delay for low-latency requirements. Batch processing frameworks like Apache Spark are suitable for tasks where latency is not a critical concern.

Factor 4 – Integration and output

This table highlights key factors for deciding the output format, whether it’s for databases, Business Intelligence tools, or custom systems. By using compatible technologies, organizations can ensure the transformed data fits seamlessly where it’s needed.

CharacteristicConsiderationCriteria
Output FormatDefine the desired output format — relational databases, Business Intelligence tools, or custom downstream systems.Choose technologies compatible with the output format. SQL-based solutions are suitable for relational datasets, while stream processing frameworks like Apache Kafka can feed real-time data into downstream systems. Big Data tools can store data in column-based file formats such as Parquet.
Integration NeedsAssess integration requirements with other systems and applications.GUI-based tools like Dataiku and KNIME offer user-friendly interfaces for seamless integration. Programming languages like Python and Java provide extensive libraries, APIs and SDKs, facilitating more robust integration with various systems.

Factor 5 – User expertise and collaboration

Success in data transformations depends on both technology and teamwork. Here we focus on two vital aspects: assessing team members’ skills, including analysts, engineers, and non-technical users, and understanding collaboration between technical and non-technical stakeholders.

CharacteristicConsiderationCriteria
Team Skill SetEvaluate the proficiency of team members—data analysts, engineers, non-technical users.Choose technologies aligning with the team’s expertise, whether it’s for Data Analysts, Software Engineers, or non-technical users. For non-technical users, intuitive GUI-based tools like DataIku provide accessible interfaces. For technical users, programming languages and frameworks offer flexibility and customisation options. The majority of analysts possess proficiency in SQL, hence, they tend to favour this solution when they act as end users.
Collaboration NeedsConsider collaboration requirements between technical and non-technical stakeholders.GUI-based tools promote collaboration through user-friendly interfaces and visual workflows. Collaborative features in tools like Dataiku and KNIME enable teamwork. For code-based solutions, version control systems like Git enhance collaboration among developers. Usually, when data is processed using custom software, supplementary tools like Amazon Athena or Looker are often required for presenting the data.”

Factor 6 – Scalability and performance

Let’s focus on two important aspects: ensuring the chosen technology handles growing data and variable workloads. Also, make sure to use optimized tools for large-scale transformations.

CharacteristicConsiderationCriteria
Data SecurityPrioritise data security, especially for sensitive information.Choose technologies with robust security features, including encryption at rest and in transit. Cloud-based platforms often provide built-in security measures. Implement access controls and encryption for sensitive data storage.
Compliance RequirementsAdhere to industry regulations and compliance standards.Select technologies compliant with necessary certifications such as HIPAA, GDPR, or PCI DSS. Cloud providers often offer compliant solutions. Implement auditing and monitoring tools to ensure compliance with regulations.

Factor 7 – Security and compliance

In today’s data-driven world, protecting data and following regulations are essential. By choosing secure technologies with encryption and access controls, organizations can protect their data.

CharacteristicConsiderationCriteria
Data SecurityPrioritise data security, especially for sensitive information.Choose technologies with robust security features, including encryption at rest and in transit. Cloud-based platforms often provide built-in security measures. Implement access controls and encryption for sensitive data storage.
Compliance RequirementsAdhere to industry regulations and compliance standards.Select technologies compliant with necessary certifications such as HIPAA, GDPR, or PCI DSS. Cloud providers often offer compliant solutions. Implement auditing and monitoring tools to ensure compliance with regulations.

Factor 8 – Cost and resource considerations

The availability of skilled professionals, training, and support is crucial. Opting for technologies with active communities and robust support ensures continuous learning and assistance. This support and continuous learning is often found with software engineering partners.

CharacteristicConsiderationCriteria
Infrastructure CostsEvaluate the cost implications of the chosen technology.Cloud-based solutions offer cost-effective scalability, allowing businesses to pay for resources as needed. However, the base cost of running transformations in the cloud is typically higher than on-premise solutions. Solutions that enable local development and testing will be much cheaper than their cloud-based counterparts. It’s important to consider the total cost of ownership (TCO) over time, which includes licensing fees, maintenance, and support costs for on-premises solutions.
Resource AvailabilityConsider the availability of resources such as skilled professionals, training programs, and support services.Choose technologies with ample resources and community support. Active communities provide forums, tutorials, and resources, ensuring continuous learning and problem-solving. Prioritise platforms with robust support services, ensuring prompt assistance when issues arise.
Open-source vs prioprietary solutionConsider community support to easily extend the solution and quickly find bugs.Open source solutions often offer greater flexibility, enhanced security through transparent code, and a collaborative community for innovation, as opposed to proprietary alternatives, which can be restrictive and less adaptable to specific user needs.

Factor 9 – Type and lifecycle of a project

Deciding on the type and size of a project is essential in guiding our choice of technology to fulfill our objectives.

CharacteristicConsiderationCriteria
Project TypeDetermine if it’s a research prototype or a long-term, critical, production-grade system.In a research prototype, choosing functionality over high optimisation and generic code is acceptable. Conversely, for a production-grade system, prioritising tools that encourage generalisation and robust software engineering practices is favourable.
Size of the ProjectAssess the size of the project, distinguishing between small-scale and large-scale projects.Small projects may not demand generalisation, focusing on code readability for easier comprehension. For large-scale projects, emphasise readable, maintainable code, supporting software engineering practices.

Navigating the data transformation landscape

As we reflect on the journey through various data transformation technologies, it becomes evident that their role is pivotal in shaping the narrative of modern analytics. Let’s recap the essence of our exploration and underscore the vital aspects that demand attention.

Data transformations serve as the catalyst, transforming raw, unwieldy data into refined, structured formats ready for analysis. They enable organizations to extract meaningful patterns, identify trends, and make informed decisions. By cleansing, aggregating, and enriching data, transformations empower businesses to unearth valuable insights hidden within complex datasets.

The significance of choosing the right technology

Matching specific requirements and considerations with the right tool is similar to finding the perfect key for a complex lock. Tailoring technology choices to the nuances of the organization and the data—whether structured or unstructured, massive or modest—ensures optimal performance, scalability, and cost-efficiency. 

A thoughtful selection process guarantees that the chosen solution aligns seamlessly with the organization’s goals and resources. Since every technology has a learning curve, or you might need to find an expert in the field of data engineering, you might also consider working with a software engineering partner.

Experienced professionals bring innovative perspectives, technical expertise, and efficiency to your data transformation endeavors. This allows you to leverage cutting-edge technologies and industry best practices, ensuring your projects are successful and future-proofed against emerging challenges.

Automation

Automation in data transformation is a pivotal aspect of revolutionizing the efficiency and accuracy of data processing workflows. By leveraging automated processes and tools, organizations streamline repetitive tasks such as data extraction, cleansing, transformation, and loading. 

Automation expedites the handling of large volumes of data and minimizes human error, ensuring consistency and reliability in the transformed datasets. Additionally, automation paves the way for enhanced scalability, enabling systems to adapt to evolving data requirements seamlessly. Overall, automation in data transformation empowers businesses to focus on strategic decision-making by accelerating processes and ensuring data accuracy.

The future of data transformations

As we glimpse the future, we anticipate an exciting evolution in data transformations. Emerging technologies such as AI-driven automation, advanced machine learning algorithms, and real-time processing frameworks will redefine the landscape. 

Augmented analytics, powered by natural language processing and automated insights generation, promises to revolutionize how we interact with data. By staying abreast of these trends and embracing innovative tools, organizations can prepare for a future where data transformations seamlessly adapt to the ever-changing demands of the digital age.

The possibilities are limitless with the right technology, mindset, and strategic vision. So, explore the depths of your data, and let the artistry of data transformations illuminate your path to success.

1. What is data transformation and why is it crucial for data analysts?

Data transformation involves converting raw data into a structured format suitable for analysis. It includes cleaning, aggregating, and manipulating data to derive meaningful insights. It’s crucial for data analysts as it ensures accuracy, consistency, and relevance in the data used for decision-making.

2. How does data transformation enhance analytical processes?

Data transformation streamlines and standardizes data, making it more accessible and understandable. It facilitates easier identifying trends, patterns, and outliers, enabling data analysts to draw accurate conclusions and make informed business decisions.

3. What techniques are employed in data transformation?

Data analysts use techniques like data cleaning to remove errors, data integration to combine different datasets, data normalization to standardize data formats, and data aggregation to summarise information. They may also use tools like ETL (Extract, Transform, Load) processes or scripting languages like Python or R. In some cases, Analyst can transform data on their own. However, to perform complex data transformations, a data engineer is required.

4. How does data transformation impact business insights?

Data analysts can uncover correlations, identify trends, and predict future outcomes by transforming raw data into a usable format. This empowers businesses to make data-driven decisions, improve operational efficiencies, optimize strategies, and enhance overall performance.

5. What are the benefits of automating data transformation processes?

Automation in data transformation reduces manual effort, minimizes errors, and accelerates the speed of analysis. It allows analysts to focus more on interpreting results and strategizing, leading to quicker insights and better decision-making.

Curated by

Sebastian Synowiec

Liked the article?

Share it with others!

explore more on

Take the first step to a sustained competitive edge for your business

Let's connect

VirtusLab's work has met the mark several times over, and their latest project is no exception. The team is efficient, hard-working, and trustworthy. Customers can expect a proactive team that drives results.

Stephen Rooke
Stephen RookeDirector of Software Development @ Extreme Reach

VirtusLab's engineers are truly Strapi extensions experts. Their knowledge and expertise in the area of Strapi plugins gave us the opportunity to lift our multi-brand CMS implementation to a different level.

facile logo
Leonardo PoddaEngineering Manager @ Facile.it

VirtusLab has been an incredible partner since the early development of Scala 3, essential to a mature and stable Scala 3 ecosystem.

Martin_Odersky
Martin OderskyHead of Programming Research Group @ EPFL

VirtusLab's strength is its knowledge of the latest trends and technologies for creating UIs and its ability to design complex applications. The VirtusLab team's in-depth knowledge, understanding, and experience of MIS systems have been invaluable to us in developing our product. The team is professional and delivers on time – we greatly appreciated this efficiency when working with them.

Michael_Grant
Michael GrantDirector of Development @ Cyber Sec Company

VirtusLab has built an excellent platform with exceptional development talent and a strong customer base.

vincenzo_la_ruffa
Vincenzo La RuffaPresident @ Aquiline Capital Partners