Background
Our client is an Austrian chemical company and is one of the world’s largest producer of polyethylene (PE) and polypropylene (PP), with 6.900 employees operating over 120 countries. The company is known for its production and distribution of polyolefin, base chemical and fertilizer solutions.
Challenge
In today's competitive and complex market, reacting swiftly to set product prices is important due to the unpredictable price fluctuations in raw material markets volatility. To overcome these challenges effectively, a deep understanding of market dynamics is necessary, along with the ability to extract valuable insights from complex and diverse datasets. Additionally, proactive risk mitigation strategies are essential to address supply chain disruptions, market uncertainties, and unexpected events impacting raw material prices. These aspects demand a comprehensive approach to pricing decisions in this fluctuating market landscape.
Approach
By collecting as many data sources as possible (internal or/and external), create a dataset that will benefit from AI techniques. Feed that dataset to Tangent Work’s state-of-the-art timeseries predictive engine and select best model from a multitude of automatically trained models. Translate the results of the algorithm and make them available into an intuitive application, allowing the client to visualize the forecasted prices and main market drivers so that they can adapt their pricing process.
Key results
Benefits
By implementing a state-of-the-art AI model, our client was able to gain insights and identify market drivers that were unknown thus far. The solution also showed a high predictive potential in both short-term (1 month) and mid-term (3 months) prices, allowing them to anticipate and take advantage of this information in their pricing decision. It also offered our client a tool to perform root cause analysis on historical data, which gave them additional knowledge on the those markets.
Technology partners
Tangent Works
Tangent Works specializes in automated predictive model-building engines. They automate forecasting and anomaly detection using time series data analysis, generating accurate models based on detected patterns.
Share this key play
