The Trust Problem at the Heart of AI-Driven Investment Analysis

Mike Ryan, Founder and CEO, Bullet Point Network

Mike Ryan is the Founder and CEO of Bullet Point Network. He previously served as Partner & Global Head of Equities at Goldman Sachs and as an Investment Committee Member and Head of Equities and Absolute Return at the Harvard Management Company.

Institutional rigor in an AI-Accelerated world

In today's private markets, data has never been more abundant. Research tools have never been more accessible. AI has never been more ubiquitous. And yet, for most venture capital and private equity firms — and the limited partners and direct investors who coinvest alongside them — confident decision-making cannot arise from simple information gathering or generic summaries generated in a few minutes.

To drive high-conviction conclusions, you need to trust the answers you're getting use data to stresstest your assumptions, quantify upside and downside risk, and seamlessly connect your thesis to the evidence and calculations that actually matter. Once you can, in fact, produce customized, decision-grade materials in 80 percent less time, your team can focus on critical thinking, primary research, and informed debate — or move on to new idea generation, sourcing, or value-add work with existing portfolio companies.

That tension — between the speed AI delivers and the confidence serious decisions demand — is precisely the problem Bullet Point Network was built to solve. After two decades of doing serious investment analysis at Goldman Sachs and the Harvard Endowment, and two years of struggling with generic AI tools, Mike Ryan saw the need to develop a comprehensive platform that transforms data analysis, fundamental research insights, and nuanced judgment into institutional-quality analysis with the depth and precision serious investors require.

What Real Investment Work Actually Requires

In theory, AI tools search all available information and produce outputs tailored to each investment decision. In practice, unreliable sources, context window limitations, simplistic calculations, and outright hallucinations often produce generic or dangerously misleading work. AI workflows need to mirror and extend what serious investors actually do every day: configuring each memo or slide deck to cover exactly what the thesis demands for each unique company, prioritizing sources considered most relevant and reliable for each topic, driving calculations from a spreadsheet they know and trust, stress-testing assumptions, building logical scenarios, adding nuance to conclusions, and allowing for fluid iteration and team debate before reaching a final decision.

  ​BPN has been a strategic partner, enabling us to quickly frame detailed analyses and summarize various possible outcomes with data to support our assumptions.  

Formats vary by fund, investment and use case. Each company is unique, and the thesis drives which topics need the deepest research. That is the real work — and it needs to meet an institutional standard before speed and scaling efficiency become the conversation.

Configurability and Source Control as Core Design Principles

BPN's AI+1 platform is architected around a principle most AI tools treat as an afterthought: full configurability and customization with fluid iteration based on transparency and control. Every analysis is built from a firm's own validated source documents, proprietary call transcripts, and familiar spreadsheet models, alongside data from trusted providers and the public internet. A proprietary multi-level prompt engineering framework and an "under the hood" AI agent — trained to work like an investment analyst — ensure the platform is not producing generic summaries or unsupported claims. BPN's proprietary sourcepairing prioritizes the most relevant information and its LLM model selection routes queries to the most appropriate model for each topic. The result is answers that are traceable, auditable, and grounded in sources your team can trust.

From Sources to Spreadsheet to Slide Deck

The investment industry is moving rapidly toward AI-augmented workflows. The largest LLMs are adding reasoning models, skills automation, enterprise search, and data connections. Point solutions exist across each of these verticals. BPN separates itself from point solutions because its fully integrated workflow ensures consistency, fluidity and contextual relevance. Sources are vetted, tagged, and paired to each prompt. Analysts connect their own spreadsheet — with the inputs, formulas, and model architecture that frames the analysis the way they want to see it — and the platform enables them to stresstest assumptions, build scenarios, and produce Tableau-like charts to embed in memos and slides. Crucially, the story and the numbers stay connected, and conclusions are based on source documents, trained investment logic, and trustworthy calculations — not AI guesses.

The Emerging Standard for Investment Intelligence

Gathering and summarizing information is easier than ever. The real constraint is using AI to maximize insight and drive better decisions, which requires seamless integration between AI workflows, trusted data sources, customizable output and fluid iteration.

Simply adopting AI or deploying point solutions will not create an advantage; In fact, it may reduce rigor and waste time. On the other hand, firms that deploy solutions specifically designed for the complexity, precision, and rigor of institutional investment decision making, and who leverage their genuine domain expertise, will achieve a rare combination of material time savings and simultaneous quality improvement. BPN's AI+1 Platform represents that standard: a fully integrated, configurable, and traceable solution that arms investment teams to think better, move faster, and communicate with the clarity that critical investment decision-making requires.