The popularity of analytics engineers grows every day. It creates additional confusion for some data leaders as they don’t understand if they should choose data analysts, data engineers, data scientists or someone else. I would like to clarify it as I have more than 8 years of experience in data across different functions and come from a statistics background. I am the data ninja combining everything. My business acumen skills, experience as an analyst on streaming and batch data, and high technical knowledge of SQL, DBT data modelling, and building data pipelines make me a perfect analytics engineer. In my everyday work, I ensure that data is transformed, structured, and made accessible to less technical analysts and stakeholders so that they can drive business decisions. But what skills should a good analytics engineer have?
1. Strong SQL Skills
SQL is the foundation of an analytics engineer’s toolkit. A good analytics engineer should be proficient in writing efficient and optimised queries, handling complex joins aggregations, and working with large datasets. My knowledge of SQL allows me to create rules in SQL, such as if this then that, to be sure that the query runs when there is data at the source, create validations for duplicates or missing data and more. My knowledge of advanced SQL techniques such as window functions, common table expressions (CTEs), and indexing strategies is powerful. I do not use SQL only to get information from data; I create the scripts that build the tables and products.
2. Data Modeling and Warehousing
Understanding data modelling concepts like star schema, snowflake schema, and normalisation is essential. I often work on data migration projects from one system to another, such as internal legacies to GCP at Inmarsat or Datarama to GCP at the7stars. My role is to know how to optimise the process using appropriate models and designs that vary depending on the business needs. My power comes not from the tool I use but from knowing how to use a specific approach and quickly pick up the needed tool. I have worked across many platforms like BigQuery or Databricks for Analytics on Cloud. Most importantly, I know when and how to use it rather than if I have a specific experience on this or another platform.
3. Proficiency in dbt (Data Build Tool)
There are different tools for ETL, and although knowledge of DBT is not always needed, it has become a standard tool for transforming raw data into structured, analytics-ready data models. Even if I do not use DBT for a specific project, the knowledge of proper documentation and how this should look in DBT allows me to transfer from one tool to another. Analytics engineers should be adept at writing and managing DBT models, macros, and documentation to maintain a well-organised data transformation pipeline.
4. Python for Data Processing
While SQL handles most transformations, Python is valuable for advanced data processing, automation, integration with APIs, and creating statistical models if needed. Knowledge of libraries like Pandas, NumPy, and Airflow can help an analytics engineer perform tasks beyond SQL. Remember that analytics engineers come pretty often from an analytics background; still, they have mastered the data engineering elements, so we not only can do data engineering but also create dashboards, perform required analytics, and take insights. We are the combination of two worlds.
5. ETL/ELT and Data Pipeline Orchestration
A good analytics engineer should understand ETL (Extract, Transform, Load) and ELT (Extract, Load, Transform) processes. For example, I help connect data from the source to BigQuery for my current client. I need to extract the data from different sources, such as API, flat file, or some other streams, load this data into the pipeline, and decide where the specific transformation will occur and where the particular metric will be created. Is it on the ETL level? BigQuery level? or the Looker level? There are nonagregable metrics that can be created only at the data level and metrics that, due to the complexity of transformation, cannot be performed in ETL tool but at the BigQuery. There are many different things to consider when building the pipeline, and analytics knowledge is handy.
6. Version Control and CI/CD
Version control with Git is essential for maintaining a clean and collaborative workflow. This comes kind of like knowledge of using Microsoft Office as the version control is everywhere and using GitHub or some other version control tool is like using a piece of paper to write a note and often people collaborate on the project to build something together.
7. Data Quality and Testing
Ensuring data accuracy and reliability is critical. The right analytics engineer should implement automated testing with tools like DBT tests or custom validation scripts to catch inconsistencies and errors early. Some bugs will trigger the error or warning, but sometimes they may not, and we need to be able to pick them up. For example, specific calculations on the column will not bring the expected effect, or the system will not provide full data in the table. We need to know when to validate on the SQL level and when to validate on the visualisation dashboard level. If the data comes through multiple different parts of the pipeline, when should it be validated, and what?
8. Business Acumen and Communication Skills
My journey in data started from being an insights analyst, and my passion for business allowed me to join the Microsoft Advertising team. I see how often my understanding of business increases because I have my own Data Science Consultancy. Clients tell me about their problems and needs, which I can easily relate to. Their problems are mine problems, too. Empathy allows me to grow
9. Cloud and Infrastructure Knowledge
Familiarity with cloud platforms such as AWS, GCP, or Azure is increasingly essential—understanding services like BigQuery and IAM permissions helps manage and secure data infrastructure. The data that I work with is the Cloud data. I work with data streaming from API or even machine data such as satellite. I need to know how to move around the Cloud, set up rules and permissions, schedule processes, and protect the data in the Cloud.
Final Thoughts
The best analytics engineers aren’t just technical experts but business-minded problem solvers. Mastering these skills can bridge the gap between raw data and actionable insights. What do you think? What other skills do you believe are essential for an analytics engineer?
