Dataloopr

Recently, a candidate went through the assessment process at ADIA (Abu Dhabi Investment Authority) for an analytics/data-focused role. The assessment was structured to test a broad mix of technical, analytical, and problem-solving skills, reflecting the kind of work expected in a real-world investment and data environment.

The assessment was divided into two main rounds:

Round 1: Online Assessment on HackerRank

Duration: 60 minutes
Format: 20 Questions

This round was fast-paced and tested both breadth and depth of knowledge. The questions were distributed across coding, mathematics, computer science, and data interpretation.

1. Coding Problems (2 questions)

• SQL Problem:
  A straightforward query involving joins and aggregations. Nothing too advanced, but you needed to be sharp on writing clean queries under time pressure.

• Python Problem:
  A moving window problem, which required iterating through sequences and applying logic efficiently. The difficulty level was easy to medium, but careful handling of indices and edge cases was essential.

2. Math Problems

The math portion leaned heavily on fundamentals. Topics included:

• Probability & Bayes’ Theorem 

• Expected Value 

• Algebra

• Math Puzzles

These were not overly complex but required quick thinking and accuracy.

3. Computer Science Concepts

There were theoretical questions testing understanding of:

• Time & Space Complexity – analyzing algorithms and recognizing Big-O behavior.

• Sorting Algorithms – understanding efficiency, edge cases, and use cases for different sorting approaches.

4. Data Interpretation

This section resembled GMAT-style questions where you’re given tables and charts and need to extract insights quickly. It tested your ability to read data carefully, spot trends, and perform quick calculations under time pressure.

👉 Takeaway: Round 1 was less about deep complexity and more about speed, accuracy, and clarity of fundamentals.

Round 2: HackerRank - Data Analysis / Pipeline Building

Duration: 4 hours
Format: 1 Comprehensive Question

This was a practical, hands-on assignment where you are given a dataset and asked to simulate a real-world data workflow. The task included:

1. Finding anomalies in the dataset.

2. Cleaning the data and preparing it for analysis.

3. Saving the cleaned dataset into a structured table.

4. Writing SQL queries on top of the cleaned data.

5. Enhancing a signal from the data (e.g., creating a feature or deriving a meaningful metric).

6. Writing clean, modular code to ensure readability and maintainability.

7. Building a pipeline that could automate this workflow end-to-end.

This round tested not just coding skills but also:

• Data intuition (spotting irregularities).

• Best practices in coding (modularity, documentation).

• Practical SQL fluency.

• Ability to design scalable pipelines rather than one-off scripts.

👉 Takeaway: Round 2 was designed to mirror day-to-day responsibilities of working with large datasets in investment or analytics settings. It tested both technical execution and engineering rigor.

Final Thoughts

The ADIA assessment was a comprehensive test of skills across multiple dimensions:

• Quick thinking (Round 1).

• Real-world data handling and engineering (Round 2).

It stood out from typical assessments because it balanced theoretical knowledge, coding ability, and applied problem-solving. For anyone preparing, I would recommend:

• Practicing SQL joins and aggregations.

• Brushing up on probability and Bayes’ theorem.

• Reviewing algorithm complexity basics.

• Practicing quick GMAT-style data interpretation sets.

• Building small end-to-end data pipelines with clean code.

Related Articles

Quant Research Interview Preparation: ADIA, Qube Research, ADS

Quant Finance Basics - Market Making Optimization and Execution

Quant Analyst Interview Questions at Millennium