Automated data scraping platform powered by AI and LLMs

Rothwand Case Study AI Data LLMs

Collecting data from thousands of sites using AI and LLMs

What does Rotwand do?

Rotwand is a boutique PR agency in Munich, Germany, that focuses on data-driven public relations. The agency combines traditional PR methods with advanced SEO techniques to create effective digital PR solutions. This approach increases visibility, generates leads, and provides measurable results for their clients.

Rotwand has been featured in notable publications such as PRWeek, The Holmes Report, Handelsblatt, ARD, Frankfurter Allgemeine Zeitung, BR, WIRED, and HORIZONT. Rotwand, an independent and progressive company, wants to become a leader in the high-tech PR industry. The company improves public relations with creative methods and strategic insights. To stay competitive, Rotwand is dedicated to using AI to refine its approach and offer better solutions for its clients.

How does Vstorm cooperate with Rotwand?

A few years ago, Rotwand started a project using traditional data scraping techniques to collect unstructured data from many different sources. This approach required a lot of budget, time, and development work. Additionally, the accuracy of data collection was unsatisfactory.

Rotwand faced a big challenge: how to efficiently and accurately scrape unstructured data from numerous sources while minimizing costs and development efforts. Seeing the limitations of their traditional methods, they approached us at Vstorm, as experts in AI and custom LLM-based software.

Our goal was clear: to develop a data scraping technique using advanced Natural Language Processing (NLP), Machine Learning (ML), and Large Language Models (LLMs). This approach aimed to significantly reduce development costs, increase accuracy, and minimize programming efforts. We aimed to create a single algorithm capable of understanding context and accurately extracting data from diverse sources, which is too time-consuming for a human to handle.

We designed a Hyper-Automated Platform for Rotwand, intended to scrape large amounts of unstructured data from thousands of news platforms. This platform is scalable and can be expanded over time to include more sources. The key to this solution is Information Extraction technology, which enables precise and context-aware data collection.

  • We built the information extraction solution using advanced natural language processing models and machine learning algorithms. For data scraping, we utilized Playwright to efficiently gather data from multiple news platforms. The scraped data is then processed and prepared for analysis
  • To hand.le our machine learning needs and manage complex language model operations, we incorporated LangChain technology along with Pydantic for data validation. Python was our primary development language, and we employed various cloud services for data processing and storage.
  • The system runs weekly using Celery Beat in Python, with logs managed by a Redis broker to ensure smooth operation and timely updates. Throughout the project, we ensured that all technologies and software complied with copyright laws and intellectual property rights, guaranteeing ethical use and distribution.


By implementing an information extraction solution, Rotwand significantly improved its media monitoring. The system automated the gathering of news data, reducing the time and labor required for manual monitoring, and efficiently processed hundreds of thousands of articles, scaling to handle millions annually.

The solution provided weekly updates, allowing Rotwand to quickly adapt to emerging information. Leveraging large language models (LLMs), and AI the system enhanced accuracy by understanding the context and sentiment of scraped unstructured data, providing nuanced insights into media presence.

The highly scalable hyper-automation platform allowed Rotwand to grow with its expanding needs, ensuring it stayed competitive and made informed decisions in the public relations industry.