Scalable and Adaptive System to Infer Industry Sectors of Companies: Prompt + Model Tuning of Generative Language Models – Lele Cao & Vilhelm von Ehrenheim, EQT

Many Private Equity (PE) firms, like EQT, focus on themes when making investments. They spot industry trends and invest in promising companies within those sectors.

Session Outline

Many Private Equity (PE) firms, like EQT, focus on themes when making investments. This means they try to spot trends across various industry sectors and invest in promising companies within those sectors. Identifying the right sectors for these companies is key to the success of thematic PE funds. EQT’s Motherbrain will present its solution, which uses prompt engineering and model tuning of generative language models to help with this process.

Key Takeaways

  • We standardize the sector framework and discuss the typical challenges.
  • We introduce our sector inference system addressing these challenges. Our system is built on a medium-sized generative language model, finetuned with a prompt + model tuning procedure.
  • The deployed model demonstrates a superior performance than the common baselines.
  • The system has been serving many investment professionals for over a year, showing great scalability to data volume and adaptability to any change in sector framework and/or annotation.
Add a comment

Leave a Reply