Transforming deal execution with Generative AI

As interest rates drop and high-quality assets flow into the market, Private Equity (PE) organizations across Canada are experiencing increased activity that hasn’t been seen since 2022. Alongside that spike in activity, the deal environment is becoming increasingly competitive. Firms and managers are under pressure to step up the pace and get deals finalized faster.

Part of that acceleration means unpacking and understanding the value drivers and the quality of assets. Most asset managers know this to be taxing and time-consuming work. Due diligence, data analysis and decision-making aren’t “speedy” activities. It takes time and effort to gather financial and sector information, develop lists of diligence questions, process and classify that knowledge, summarize it, make informed decisions, write deal memos, produce reports, and so on. As that list of tasks extends, so does the deal timetable.

Fortunately for investment professionals, Generative Artificial Intelligence (GenAI) tools such as Kleo, developed by KPMG, can help them work more productively and efficiently. Remember when digital spreadsheets were introduced in accounting? There’s a clear parallel to observe with GenAI. Both are computer applications that can significantly reduce the time taken to complete tasks. Digital spreadsheets didn’t reduce the number of accountants. Rather, accountants are now outsourcing number-crunching tasks to machines to do their jobs better.

PE firms now have an opportunity to harness GenAI in a sophisticated and strategic way. Identifying use cases that are ripe for automation and using GenAI the right way can improve deal flow, support decision-making, and significantly reduce costs in the months ahead.

For that, investment professionals need a deeper understanding of what makes GenAI tick. GenAI tools use machine learning, pattern recognition, and statistical analysis to process data and generate new content. A tool like KPMG’s Kleo, after training on a Large Language Model (LLM)) and can respond to user prompts for specific information in natural language and generate text responses in human-sounding language with remarkable accuracy.

In 2022, when GenAI burst onto the scene with its first “killer app,” people took to the service instantly.

• 1 million users in just 5 days¹
• 180.5 million users monthly as of July 2024²
• GenAI is currently pervasive in day-to-day activities, used primarily for text-related tasks³
• 40% use GenAI to provide document summaries. Other top use cases include document processing and knowledge management (17%), content generation (14%), and report generation (5%)⁴

That instant enthusiasm and monumental growth is a clear demonstration of GenAI’s value. Experiments and explorations might have gotten the ball rolling, but early adopters are now experts and are leveraging their expertise to accelerate and improve organizational workflow.

PE firms should seize the moment and follow suit. They can harness GenAI in the immediate term to achieve several “quick win” process optimizations, followed by broader and more targeted workflow integrations in the medium-term. Over the longer-term, GenAI efforts will have a cumulative and transformative effect. PE firms who engage these tools strategically have the potential to build a knowledge base so unique to their organization it fulfills the function of an experienced assistant.

Leveraging the capabilities of GenAI, we consider three approaches investment professionals can take in order to work smarter and more efficiently.

These tasks are highly relevant for the PE sector. In fact, many investment professionals are already leveraging GenAI tools to enhance quality, efficiency, save time, and reduce costs. Firms now have the potential to learn from their associates and turn those informal applications into broader implementations, making them part of day-to-day workflow across the organization. Opportunity areas to tackle today include:

• Text generation support for transaction teams: Use GenAI to develop customized writing guide templates and other internal resources to help newer staff generate deal memos, reports, and other documents that follow tone, style, formatting and branding guidelines. A GenAI bot can assist in drafting major components of finalized reports, investment summaries, presentations, and other content from minimal, unstructured text input.

• Meeting recording and transcription: Eliminating manual notes at meetings delivers a 100% efficiency gain. Instead, use AI software to record and transcribe meetings, then run prompts against those transcriptions in your GPT to generate other content you need, such as outlines, summaries, action items, and reports.

These quick win activities familiarize teams with prompt engineering (the intelligent crafting of queries and inputs into GenAI software to produce high quality outputs). Even better, they can deliver immediate ROI.

The next step is to develop workflow augmentations that apply GenAI across the investment lifecycle. PE firms use various resources to process financial data, but new integrations can help synthesize, automate, and accelerate that work at home. Ultimately, GenAI will put an end to the manual performance of deal-side document processing.

At the sourcing stage, GenAI can help with news analysis, thesis testing, and conference support. During diligence, it can optimize data gathering and diligence preparation, financial analysis, and report writing. As for value creation, GenAI tools can be instrumental in strategy and initiative development.

For example, some GenAI tools are being developed to parse and summarize all the qualitative content in a virtual data room. This eventual GenAI implementation will help dealmakers and analysts grasp essential information without spending time wading through excessive text. GenAI’s context-aware search function helps zero in on comparable financial information from multiple internal and external datasets quickly or list out references to compliance or regulatory issues across a range of text sources.

GenAI use cases in diligence include:

• Generating diligence question lists for potential acquisitions or for specific sectors.
• Analyzing past diligence processes or lists to generate relevant questions or identify gaps.
• Sorting and classifying financial statements, contractual agreements, and compliance documentation automatically for easier review.
   

In financial analysis, GenAI supports:

• Modeling and analysis: For instance, an analyst can use GenAI to assist in creating and incorporating formulas in Excel based on specific criteria to help develop a complex revenue forecasting model for a target company that incorporates key variables (e.g., market growth rate, seasonal factors, historical sales data).
• Summarizing management interviews according to specific criteria (e.g., strategic changes, drivers of financial trends, management outlook or to generate Q&A).

• Interpretative assistance: Get instant context for regulations and details that underlie transactions (e.g. Ask GenAI to provide examples and interpretations of relevant GAAP standards, compare accounting methods (accrual vs. cash), etc.).
   

GenAI expedites and improves report writing:

• Synthesize data from multiple sources to draft sections of a report (e.g. the major trends of an income statement).

• Review reports for terms (abbreviations, industry jargon) which may require a glossary definition, and instantly create those glossaries.
• Generate and incorporate data visualizations into reports.
   

Case study:

At KPMG, our teams have observed first-hand what GenAI can do for workflow process optimization. We’ve developed and applied GenAI accelerators to the document review process, financial data extraction, client service operations, and value creation in legal due diligence. These efforts have garnered us significant time and cost savings.

KPMG in the US used GenAI to speed up a label creation project, for instance, and reduced the launch cycle from 6-10 weeks to just 5-6 days. The project required labels for various product types for various countries in multiple languages, according to a range of design and regulatory requirements. With GenAI helping us to ingest structured and unstructured design and compliance information with remarkable efficiency, we eliminated not just hours, but weeks of manual work.

Over the longer term, there’s potential for GenAI to become more deeply integrated into work practices. As PE firms integrate medium-term optimizations across the deal lifecycle, they’re shaping a knowledge base on their deal history, systems, and processes. In the near future, PE firms will build highly knowledgeable, ultra-efficient, custom-built “bot” assistants on this foundation, designed to support human counterparts through the deal lifecycle in intelligent, efficient, and cost-effective ways.

These kinds of GPTs are currently being developed on Chartered Professional Accountant (CPA) foundations to perform accountancy tasks. We envision a team of similarly trained virtual assistants emerging in the PE space.

In the meantime, PE firms must decide what GenAI tools to use, in what areas, and whether to focus a longer-term strategy on building their own GPT. The availability of GenAI tools is proliferating, but also evolving rapidly. Investment managers need to educate themselves about publicly available solutions, how those services fit their needs, and determine whether they would benefit even more by building their own customized solution.

Data privacy and confidentiality are always a concern, and for PE professionals, they remain roadblocks for some applications of GenAI. Investors need to know their information is protected and that private data hasn’t been entered into an open-source model.

There are ways to access your own private instance of GenAI tools, such as through the premium versions of online services. What’s critical is embarking on the GenAI journey from a solid foundation of information security and privacy. Individuals responsible for GenAI software procurement at your firm need to understand what’s available on the market and how it’s evolving.

PE firms possess a lot of specialized, first party data. Forward-looking investment organizations can leverage this data to build differentiated offerings. The secret sauce in GenAI, after all, is its training data. A specialized model trained on your deal history and transactions has the potential to drive value across your deals by generating highly valuable, differentiated knowledge based on your prompts on your unique knowledge base.

To leverage this opportunity, high quality training data is essential. As with any data and analytics application, GenAI is susceptible to the Garbage In, Garbage Out (GIGO) concept. To develop a differentiated GenAI use case, you need to train your underlying GPT on clean, organized, well-sourced data.

A potential issue with GenAI is data hallucination – the generation of inaccurate responses.  Understanding these limitations and having safeguards in place is essential. Employees need training in best practices for prompting language bots and obtaining desired outputs. They need to know that asking the language bot for a direct quote can help ensure a factual answer, one taken directly from the dataset.

Realizing the full value of GenAI begins with a few targeted use cases and grows from there. Whether you’re just starting out or already making use of GenAI tools, the following guideposts can help chart your journey:

• Establish a foundation for GenAI implementation that is rooted in security and privacy.
• Identify new use cases for GenAI by using it (and encouraging use) alongside your daily work.
  • Keep it open in the background.
  • Use it to review email drafts and analyses.
  • Use it to accelerate and optimize research and discovery.
  • Use it to brainstorm and stimulate your thinking.
  • Use it to identify gaps and generate new questions.
• Build awareness around GenAI and develop best practices in the workplace.
• Get your data in order to ensure it’s clean, sourced, and high-quality.
• Understand your business levers for growth.

KPMG is at an advanced stage of its own GenAI journey, with top-down support and mandatory training for employees. We’ve developed our own use cases and maintain a private instance of GenAI – a custom bot that we encourage our team members to use. With that wealth of experience, our consultants have the knowledge and capability to get PE organizations launched on their own GenAI journeys and help guide them through a long-term strategic process that makes optimal use of GenAI to save time, reduce expenses, and create value.

To learn more about where GenAI is now and where it’s headed in Private Equity, or for assistance getting started with a specific GenAI implementation or instance at your firm, contact our team.

^{1 Number of ChatGPT Users, Exploding topics, 2024}

^{2 ChatGPT Statistics (AUG 2024) – Users Growth Data, Demand Sage, August 10, 2024}

^{3 Gartner Survey Finds Generative AI Is Now the Most Frequently Deployed AI Solution in Organizations, Gartner, May 7, 2024}

^{4 Generative AI use cases could boost document and content management software, S&P Global, September 13, 2023}