Inside Scout24

Updated: March 26, 2024Reading time: 10 minutes

Data Science: A Typical Day in the World of Artificial Intelligence at Scout24 with Nandhinee Gumma

Nandhinee embarked on her journey in the field of data analysis as a working student in 2020. Following the completion of her master's degree, she transitioned into a full-time role as a Junior Data Scientist. Over time, Nandhinee's dedication and expertise have propelled her to her current position as a Professional Data Scientist. In addition to her professional accomplishments, Nandhinee is a polyglot, proficient in six languages, which enhances her ability to collaborate and communicate effectively across diverse projects and teams. Dive into the interview to uncover more about Nandhinee's fascinating journey and insights!

Can you describe a typical day as a data scientist at Scout24? What tasks and projects do you typically work on? 

As a data scientist at Scout24, my day typically involves coordinating with the team to discuss ongoing projects and prioritizing tasks. Depending on the project, I focus on data exploration and analysis, followed by model development and evaluation for projects such as real estate price prediction or fraud detection. To bring such projects to life, I participate in collaboration meetings with cross-functional teams to design innovative solutions and features, and collaborate on deploying and integrating machine learning models into production environments to enhance the platform's functionality and user experience. 

Throughout the day, there may be ad-hoc tasks, such as troubleshooting technical issues, responding to urgent requests. Flexibility and adaptability are essential as priorities may shift based on project deadlines, stakeholder feedback, or emerging opportunities. Overall, the day is a mix of technical tasks, collaboration with colleagues, and strategic planning to drive impactful outcomes for Scout24 and its users.  

How does the work of a data scientist contribute to the overall goals and strategies of Scout24? 

Projects that the data science team is involved in have wide ranging impact across various verticals. By leveraging the capabilities of data science and machine learning, we are able toanalyse user behaviour, preferences and interactions with the platform. This enables us to recommend relevant listings, tailor search results, detect and prevent any malicious activity. Ultimately, leading to higher user engagement, improved user satisfaction. Recently, we areexperimenting the use of GenAI to enhance user experience. All these initiatives align with Scout24's goal of providing a seamless and user-friendly experience for its customers. 

What are some of the unique challenges you face as a data scientist at Scout24, and how do you overcome them? 

While real estate data can often come from various sources, making it heterogeneous, incomplete and noisy, the data engineers do an excellent job at standardising the data. The data is preprocessed and cleaned, implementing data quality checks and robust data pipelines to provide availability of quality data.

Complex machine learning models may lack interpretability due to their black-box nature, making it difficult to gain insights into model decisions. To address this challenge, I employ explainable AI techniques such as feature importance analysis, model, partial dependence plots to elucidate model behaviour and make informed decisions. This also provides the stakeholders factors to understand the model behaviour.

„It is also rewarding to see the direct impact of my work on the platforms functionality, user experience and business outcomes.“

Can you share an example of a recent project you worked on that had a significant impact on Scout24's operations or products?  

As we are currently exploring the potential of GenAI and the application of LLMs in our ecosystem, I was involved in a collaborative effort with the Search team to leverage this emerging technology to enhance the search experience on our platform. The initiative aims to help users search for their desired property on our website in an easy and intuitive manner. AI-Search was introduced allowing users to describe their real estate needs in simple plain language, thereby bypassing the need to navigate through filters and dropdown menus.. This initiative positions Scout ahead of the competition in levering the power of GenAI 

How collaborative is the environment for data scientists at Scout24? Do you often work with other teams or departments? 

The environment for data scientists at Scout24 is highly collaborative, with frequent interactions and collaborations with other teams and departments being a common practice. Data scientists often work closely with cross-functional teams, including product managers, engineers, designers, and business analysts. Collaborative efforts are essential for understanding project requirements, aligning on objectives, and designing solutions that meet the needs of both users and stakeholders. 

What tools and technologies do you use on a daily basis as a data scientist at Scout24, and how do they facilitate your work? 

As a data scientist at Scout24, I utilize a variety of tools and technologies on a daily basis to perform tasks ranging from data analysis and modeling to deployment and automation. Python is the primary programming language used for data science tasks at Scout24. Python's simplicity, versatility, and extensive ecosystem make it ideal for prototyping, experimentation, and productionizing machine learning models. AWS is the chosen cloud service provider here and is used widely for data storage, computation, building, training and deployment of machine learning models in production environments. During the data exploration phase of some projects, SQL and PySparkl are used for creating custom datasets for analysis and for processing and analysing large datasets respectively. While Docker is used for containerization to ensure consistency and reproducibility, Jenkins is used for continuous integration and continuous deployment CI/CD to automate model training, evaluation and deployment.

How does Scout24 support the professional development and growth of its data science team members? 

Data scientists at Scout24 have opportunities to collaborate with teams and departments across the organization, gaining exposure to different business domains and expanding their skill set. By working on cross-functional projects, data scientists can develop a broader understanding of the company's operations, customer needs, and market dynamics, enhancing their problem-solving abilities and fostering a collaborative mindset. Further, participation in conferences to explore new technologies, methodologies to broaden the expertise and skill set are encouraged.

Can you discuss any ongoing research or innovative initiatives in the field of artificial intelligence that Scout24 is currently involved in? 

Certainly! I'm currently engaged in leveraging advanced machine learning techniques to enhance marketing strategies and effectively allocate resources. The objective of this project is to maximize the impact of marketing expenditure across diverse channels within the real estate sector. Understanding the efficacy of marketing endeavors and efficiently allocating resources is crucial for Scout24 to meet its business objectives. This initiative involves extensive collaboration with the marketing team, making it a highly cross-functional project. Overall, the media-mix-modeling project enables Scout24 to make informed, data-driven decisions, optimize marketing strategies, and maximize the effectiveness of its marketing budget. By harnessing advanced analytics techniques and predictive modeling, Scout24 can continually refine its media mix strategy, ensuring its competitiveness in the dynamic real estate market 

What advice would you give to candidates interested in pursuing a career in data science at Scout24? 

Data science roles at Scout24 require proficiency in programming languages in Python, as well as knowledge of statistical analysis, machine learning algorithms, and data manipulation techniques. I would recommend focusing on developing strong technical skills through hands-on projects, online courses, and practical experience. Data science is a collaborative field, and effective communication is essential for working with cross-functional teams and stakeholders. The ability to communicate complex technical concepts in a clear and concise manner helps in stakeholder management.

Lastly, what do you find most rewarding about working as a data scientist at Scout24, and what keeps you motivated in your role? 

As a data scientist at Scout24, one of the most rewarding aspects of my role is the opportunity to work on various different projects by leveraging data science techniques and innovative solutions. The exposure to the wide array of projects within the real estate industry provides ample opportunities to hone technical skills and as well as soft skills.

Exciting news and articles