A two-hour interactive workshop designed for real estate GPs, capital raisers, investor relations professionals, and fund managers who want to build and deploy an AI agent that runs your capital raising operation—from investor research and outreach to content creation and LP communications.
This workshop is split into two halves:
Friday, June 26: 12:00pm–2:00pm ET
The session will be recorded and shared with all participants via Circle.
Participants will receive post-course access to Circle, where we'll share recordings, agent templates, prompt libraries, skills files, and tool configurations.
You'll learn how to:
You'll leave with a working AI capital markets agent, the complete architecture for how it was built, and a repeatable system you can customize for your own fund and deploy immediately.
Part One: How to Build an AI Capital Markets Agent (approx. 60 min)
This is the technical foundation. We'll build a fully functional AI agent live, piece by piece, using Claude and Anthropic's toolset—and explain what each layer does and why it matters.
The Brain: System Instructions & Custom Prompts Every agent starts with instructions that define how it thinks. We'll write the system prompt for a capital markets agent from scratch—telling it who it is, what it knows about real estate capital raising, how it should reason about investor-deal fit, and what tone and format to use in its outputs. You'll see how the quality of these instructions is the single biggest lever on agent performance, and how to iterate them over time.
The Memory: Local Files & Fund Documents An agent without context is just a generic chatbot. We'll show how to load your agent with your fund's thesis, track record, deal memos, prior investor communications, and market research—using Claude Projects and uploaded files—so it operates with deep knowledge of your specific strategy, portfolio, and voice. This is what turns "write me an email" into "write an email that sounds like our firm and references our actual deal history."
The Skills: Task-Specific Prompt Templates Skills files are reusable instructions that teach your agent how to perform specific capital markets tasks: researching an investor, scoring mandate fit, drafting a 3-email outreach sequence, writing a newsletter, preparing a pre-meeting briefing. We'll build several of these live and show how they become the agent's playbook—a library of capabilities you can trigger on demand and refine over time.
The Communication Layer: Email, Slack & Messaging An agent that can only talk to you inside a chat window has limited utility. We'll demonstrate how to connect Claude to email (via MCP server integrations), Slack, and other communication tools—so your agent can draft and stage outreach in your actual inbox, surface alerts in your team channels, and integrate into the workflows where you already operate.
The Toolbelt: API Connections to Key Software This is where the agent becomes powerful. We'll walk through connecting Claude to external platforms—Clay for investor enrichment, Google Drive for document access, your CRM for pipeline tracking, web search for real-time market intelligence—using MCP servers and API integrations. You'll see how these connections let the agent pull live data, enrich investor profiles, and take actions across your existing stack rather than operating in isolation.
Putting It All Together: The Agent Architecture We'll step back and look at the complete system: brain + memory + skills + communication + tools. How these layers interact, how to version and improve them, and what the difference is between a copilot (you ask, it helps), a teammate (it watches for signals and drafts actions for your review), and an autonomous agent (full workflows running with human oversight at key decisions).
Part Two: Deploying the Agent—A Live Capital Raising Case Study (approx. 60 min)
Now we put the agent to work. We'll run through a complete capital raising scenario using a sample case study—a GP raising a $50M fund focused on a specific real estate strategy—and show the agent executing each phase of the campaign in real time.
Setting the Stage: Meet the GP & the Raise We'll introduce our case study GP: her fund strategy, track record, target investor profile, timeline, and constraints. This mirrors the real-world conditions most emerging and mid-size sponsors face—limited capital markets headcount, no placement agent, a 6-month window to secure commitments, and a need to punch above her weight.
🔨 Live Demo 1: Investor Research & List Building Watch the agent work. We feed it a target investor segment and it builds a structured investor universe—pulling firm profiles, identifying the right contact (not the managing partner, but the person who actually reviews deal flow), surfacing recent fund closings and mandate signals, scoring each prospect for deal fit, and producing an enriched, prioritized target list. We'll show how the agent uses its web search tools, enrichment connections, and skills files to do in minutes what traditionally takes weeks of manual research.
🔨 Live Demo 2: Personalized Outreach at Scale Watch the agent draft outreach. Starting from the enriched investor profiles, the agent generates hyper-personalized email sequences for each target—not mail-merge templates with [FIRST_NAME], but outreach grounded in each investor's specific mandate, recent activity, professional background, and connection to the deal thesis. We'll show the full 3-touch sequence: personalized opener, market data follow-up, and specific meeting ask. Then we'll show how these drafts flow into email tools ready for the GP to review and send.
🔨 Live Demo 3: Content Creation & Thought Leadership Watch the agent build the media engine. We feed it the GP's fund thesis, track record, and market data, and it produces: a structured white paper outline, a first draft of the opening sections, a newsletter edition, and a week's worth of LinkedIn post concepts—all flowing from a single authoritative document. We'll show how this content funnel works: white paper → deck → newsletter → social posts → webinar material, and why building trust through content means investors have already encountered your thinking before your outreach email arrives.
🔨 Live Demo 4: Meeting Prep & Investor Relations Watch the agent support the relationship. Before a scheduled investor call, the agent produces a custom briefing: the investor's portfolio, recent public statements, likely objections, and suggested talking points tailored to the GP's specific deal. After the meeting, it drafts a personalized follow-up referencing specific conversation points. We'll show how this same capability extends to quarterly LP updates, event strategy, and warm introduction mapping.
The Scorecard: Old Playbook vs. New Playbook We'll close the case study by comparing results—time to market, cost, reply rates, meetings booked, and pipeline generated—between the traditional approach (3+ months, $30K+, 2% reply rate, zero commitments) and the AI-powered approach (~$500/month, 3 weeks to launch, 15–20% reply rates, and a real pipeline). What it takes to get started, and why the compounding gap means every month you wait, early adopters pull further ahead.
Brad Hargreaves & Paul Stanton, Thesis Driven
Will participants receive a copy of the materials? Yes. All registered participants will receive an invitation to join our private Circle community. Recordings, slides, agent templates, skills files, and setup guides will be uploaded within one week of the workshop ending.
I can't make this time—will a recording be available? Yes. All registered participants receive access to the recording and materials via Circle, even if they cannot attend live.
Do I need a Claude account to participate? A Claude Pro or Team account is recommended to follow along in real time, but not required. All system instructions, skills files, and setup documentation will be shared via Circle so you can build the agent at your own pace after the workshop.
What if I'm already using AI tools for capital raising? This workshop is designed for both beginners and intermediate users. Part 1 covers agent architecture from first principles; Part 2 demonstrates deployment against a real scenario. Even experienced AI users consistently discover new techniques in the live build and case study demos.
