#48 - Real Estate & Financial Investment AI

Link to Folder with Everything

2026.07.15.officehours.48 — Google Drive folder

Recording of the Video

Summary (by Gemini)

Summary

Entrepreneurial presentations regarding artificial intelligence tools merged with strategic shifts toward application programming interface models for monetization.

Entrepreneurial Product Showcase

Presentations highlighted automated drone technology, real estate investment analysis platforms, and political trading tracking tools. These systems utilize machine learning and predictive data to enhance investment strategies.

Application Programming Interface Strategy

Adoption of application programming interface based business models emerged as the primary strategic decision for unlocking downstream value. Selling data access enables startups to build scalable layers upon existing platforms.

Financial Agentic Artificial Intelligence

Research on agentic artificial intelligence mapped various open source tools based on adoption, maturity, and community engagement. Experts demonstrated automated due diligence pipelines that generate investment grade memos and composite scores.

Next steps

• [Kirill Gorbounov linkedin] Send Drone Catalog: Send the drone company catalog and all available options to Rahul.
• [Rahul Singh] Publish Research Report: Edit and publish the gathered research based on personal experience.
• [Rahul Singh] Open Source Skill: Upload the domain due diligence tool to a public platform.
• [Tigist Aw] Review GitHub Skills: Explore the shared software repositories on GitHub.
• [Tigist Aw] Contact on LinkedIn: Connect with the speakers on LinkedIn.
• [The group] Contribute to Knowledge Base: Join the WhatsApp group and share useful findings in the news channel.
• [Rahul Singh] Expose Knowledge Base: Make the aggregated information accessible via web or chatbot interfaces.

Details

• Meeting Purpose and Format: Rahul Singh initiates the meeting, outlining the goal of showcasing entrepreneurs who use Artificial Intelligence to build products and grow businesses (00:00:47). They emphasize the importance of presenting these stories as role models for other knowledge workers (00:02:16). The scheduled structure includes an entrepreneur presentation followed by a 10 to 15-minute segment on the state of financial agentic Artificial Intelligence presented by Rahul Singh (00:04:40).
• Participant Introductions and Networking: Attendees introduce themselves, including Noor Ul Hassan from Lahore, Tinius S from Miami, and Tigist Aw from Northern Virginia. Participants discuss their professional backgrounds and long-standing business relationships, noting the efficiency of global collaboration (00:06:55) (00:15:29). Rahul Singh acknowledges that attendees range from Washington, District of Columbia to the East Coast, creating a collaborative environment (00:15:29).
• Discussion of Drone Technology: Kirill Gorbounov linkedin discusses a defense-related project involving autonomous interceptor drones capable of detecting and kinetically destroying offensive drones. They offer to provide the catalog and options for this technology to Rahul Singh for potential networking within the defense contracting sector (00:08:52).
• Cash Flow Real Estate Platform Overview: Kirill Gorbounov linkedin explains the Cash Flow Real Estate platform, a product developed over 15 years that analyzes property investments (00:11:01) (00:20:26). The platform currently covers Virginia, Maryland, District of Columbia, New Jersey, Delaware, Pennsylvania, and West Virginia, and requires a $100,000 investment to expand coverage to the entire United States (00:11:01) (00:36:46). While currently run as a local instance, Kirill Gorbounov linkedin expresses plans to move it online (00:20:26) (00:36:46).
• Real Estate Investment Metrics and Data: Kirill Gorbounov linkedin details the methodology for evaluating property potential, using the Debt Coverage Ratio, which measures the relationship between income and expenses, such as Principal, Interest, Taxes, and Insurance (00:19:32). The dashboard utilizes polynomial regression machine learning to analyze properties, considering factors like proximity to opportunity zones to predict growth (00:21:15). They clarify that data sources are proprietary and note that while margins can be tight for some properties, other investments yield a 40% Return on Investment (00:23:13) (00:41:14).
• Market Macroeconomics and Tax Strategy: Kirill Gorbounov linkedin notes the impact of government debt, currently at $40 trillion with $200 trillion in unfunded liabilities, on inflation and property values. They discuss the benefits of real estate ownership, including depreciation to lower taxable income and the $28 million step-up basis exemption for heirs, which allows for tax-free transfer of assets (00:24:25).
• Introduction to Politic Platform: Kirill Gorbounov linkedin presents the Politic platform, which tracks financial instruments, including derivatives, equities, and futures, which total approximately $1,000 trillion globally (00:25:15). The platform provides transparency into congressional disclosures, allowing users to monitor stock purchases by politicians such as Nancy Pelosi (00:26:25). The tool also integrates data from sources like Ark Invest and Motley Fool to allow users to overlay investment strategies (00:30:28).
• Politic Platform Analytics and Capabilities: The Politic platform includes features to track trading volume, identifying that Democrats traded 6,389 times and Republicans traded 4,900 times in the previous year (00:28:14). The platform offers a leaderboard for politicians like Ro Khanna and provides historical performance data, such as a 2,254% return on Western Digital stock (00:29:22) (00:32:46). Users can access the dashboard for free or via a $15 monthly subscription, which includes access to educational content like 140 songs on financial literacy (00:32:46) (00:35:53).
• Strategic Business Expansion and Artificial Intelligence Integration: Kirill Gorbounov linkedin discusses Zenai Automation, a holding company that integrates Artificial Intelligence for digital marketing, chatbots, and twin avatar cloning, which allows for content creation in multiple languages. They mention plans to integrate voice Artificial Intelligence for 24/7 prospect inquiries and highlight that the platform connects with 100 coders to deliver custom applications (00:38:49).
• Technical Queries Regarding Real Estate Data: Tigist Aw asks how the platform calculates Return on Investment and handles data accuracy (00:41:14) (00:43:12). Kirill Gorbounov linkedin explains that calculations account for down payments, closing costs, rent, appreciation, and depreciation benefits (00:42:10). Regarding accuracy, Kirill Gorbounov linkedin mentions the use of polynomial regression machine learning and confidence intervals to provide an Artificial Intelligence score for each property analysis (00:43:12).
• Evolution of Application Programming Interface-Based Business Models: Rahul Singh observes the transition toward Application Programming Interface-based business models, suggesting that providing data access via an Application Programming Interface can create downstream value for other companies (00:45:24). They encourage Kirill Gorbounov linkedin to sell Application Programming Interface access for the Cash Flow Real Estate data to other startups, enabling them to build additional layers of value (00:44:16).
• Landscape of Financial Agentic Artificial Intelligence: Rahul Singh presents research on the current state of financial agentic Artificial Intelligence, analyzing approximately 25 open-source and commercially supported tools (00:46:38). The selection criteria prioritized maturity, engagement, and community support, rather than just using the raw output of generative models (00:49:00). Rahul Singh notes that effective solutions often combine traditional quantitative statistical methods with generative Artificial Intelligence for planning and strategy (00:50:01) (00:54:49).
• Analysis of Available Artificial Intelligence Tools: Rahul Singh maps various tools based on adoption, freshness, and engagement, highlighting platforms like Anthropic's financial services tools, OpenB, Finn GPT, and trading agents (00:52:08). They discuss GitHub metrics, such as forks and stars, to measure the viability and community backing of these projects. The presentation notes that while many tools, such as trading agents and Storm, are relatively new (approximately 2.5 years old), they offer significant potential for specialized financial tasks (00:53:23).
• Anthropic Skills and Plugin Integration: Rahul Singh discusses the Anthropic Skills repository, noting it is a standard for Claude, and mentions the availability of the Claude Co-work plugin (00:56:09). They highlight that various partners, including Standard and Poor's Global and the London Stock Exchange Group, provide skills that allow users to integrate financial data directly into their analysis workflows (00:58:45).
• AI Skills and Commands in Financial Research: Rahul Singh explains the concept of "skills" for AI agents, describing them as structured documents and tools used to guide agents through specific tasks like equity research. They clarify that "commands" function as shortcuts that allow users to direct agents to perform tasks—such as specific research on a ticker—more efficiently (01:00:09).
• OpenBB Framework: Rahul Singh introduces OpenBB as an open-source platform for financial data rather than a standalone agent skill. They emphasize its utility for quantitative analysts to gather and manage data for specific tickers, noting that it serves as a robust tool to support agent functionality (01:01:26).
• GPT Researcher: Rahul Singh highlights GPT Researcher as a long-standing tool for deep internet research, explaining that this pattern of automated crawling has become standard in many AI tools (01:02:39). The discussion notes that users can integrate services like Tavily to handle internet crawling, which is a critical component for evaluating public assets, companies, or properties (01:04:00).
• Crawl for AI: Rahul Singh presents "Crawl for AI," an open-source project with 70,000 followers that utilizes a simple intermediate scripting language. They explain that this tool allows AI to record actions—such as logging into websites or filling out forms—and then generate code to execute these tasks, potentially facilitating data retrieval from private databases that lack APIs (01:05:08).
• AI Hedge Fund and LangGraph: Rahul Singh reviews the AI Hedge Fund project, which employs a multi-agent structure mimicking famous investors like Bill Ackman and Warren Buffett, overseen by a risk manager agent (01:07:47). They note that the architecture—specifically the use of LangChain and LangGraph to manage long-term agent workflows—is a "gold standard" in open source development (01:08:59).
• Gemini Skills: Kirill Gorbounov linkedin notes that Gemini also supports skills on paid plans, which Rahul Singh acknowledges as a positive trend in the adoption of agent-based workflows. Rahul Singh recalls that Microsoft pioneered this concept early on, referring to them as "semantic functions" (01:10:06).
• FinGPT and Bean Counter: Rahul Singh discusses FinGPT, a sophisticated open-source financial model that cost $3 million to train and requires specialized expertise to operate. Additionally, they mention "Bean Counter," a tool for managing accounting entries, suggesting it could be an interesting project to test for managing accounting across multiple companies (01:11:09).
• Due Diligence Skill Demo: Rahul Singh demonstrates a custom "Finance Domain Due Diligence" skill developed in the Zed editor. This tool analyzes companies based on domain names, running a 10-phase pipeline—covering tech stack, financials, and reputation—to produce a 0-100 composite score and an investment-grade memo (01:12:27) (01:16:12).
• Due Diligence Skill Execution: Rahul Singh runs the due diligence skill on "Dia browser" to demonstrate its capability to extract footprints and assess investment viability (01:16:12) (01:20:07). The agent successfully identifies the company, researches its market signals, and ultimately reports that the company was acquired, illustrating how the tool provides rapid, automated analysis (01:18:53) (01:24:32).
• Politic Platform: Kirill Gorbounov linkedin introduces "Politic," a platform aimed at leveraging data from stock trading patterns of politicians (01:21:27). They explain that the platform will use fine-tuning and retrieval-augmented generation (RAG) to ground the model in proprietary data, thereby reducing hallucinations and providing more accurate investment guidance than general LLMs, with a subscription cost of $14.99 per month (01:22:15).
• Business Advisory Services: Rahul Singh outlines that they are currently seeking small businesses and nonprofits to advise on AI implementation, functioning as coaches to help teams integrate these tools (01:25:37). They invite participants to present their own commercial AI projects for feedback and collaboration (01:26:59).
• Community Roadmap and Events: Rahul Singh discusses the future community roadmap, which involves aggregating knowledge bases for web or MCP server access and vetting tools for commercial-grade work (01:26:59). They announce plans to resume physical events in the DC area to teach bleeding-edge AI methods while continuing to broadcast these sessions online (01:28:13).