Skip to main content

Overview: The client is a new generation private credit firm focusing on investments in the climate, health, and educational tech sectors. They partnered with Bloom AI to leverage AI and generative AI across their research and investment value chain.

Challenge: Many aspects of the investment value chain are highly manual and time-consuming. A typical investment requires screening 100-250 companies, where speed and efficiency in research are critical. Similarly, the diligence and monitoring process demands significant manual effort. Additionally, a key aspect of an investment firm is its “knowledge,” which is often lost in numerous files and emails. The private credit client aimed to adopt a cloud-first and AI-first approach to manage their research and investment process, utilizing advanced technology to enhance efficiency and differentiation.

Solution: In Phase 1, Bloom AI developed a cloud-based research platform “DataScout” for the client on AWS. The solution includes three key features:

  1. Research Database: Bloom AI implemented a cloud-based research database to gather information on target companies in a structured and automated way. This database pulls data from key sources such as Crunchbase, public sources (like websites), and other information sources. The company information is also enriched using Large Language Models (LLMs) like Perplexity.
  2. GenAI Assistant: Bloom AI implemented a Claude 3.5 Sonnet based chat assistant where partners can query specific companies or ask questions ‘on-the-go’. This reduces the need for partners to ask several ad hoc questions of analysts. For example, partners can ask, “Tell me 5 biotech companies in Boston that have raised two rounds of funding.”
  3. Intel Database: As partners receive pitch documents and have new conversations with potential target companies, the information is stored in a structured way for future analysis. For instance, partners can ask, “What was the summary of our meetings in California last month?”

Check out the demo video here.

Conclusion: The platform has proven valuable for the client in several ways. Developing first-level screening of companies based on their criteria at a lower cost compared to traditional methods.

  1. Accessing key information of target companies “on the go.”
  2. Building an institutional knowledge database that aids in deal assessment.
  3. Establishing a baseline architecture that can potentially expand to due diligence (document research), portfolio tracking (e.g., quarterly reports), and monitoring (covenant monitoring).

Leave a Reply