Unlocking Value AI Bulletin #02
AI tools, tactics & case studies for real estate.
Welcome to a new issue of the Unlocking Real Estate Value newsletter. Each week I will provide you with exclusive advice and professional insights to help you realise long-term value through real estate development.
Welcome to the second edition of the Unlocking Value AI Bulletin.
This week’s curation focuses on how AI is reshaping workflows, redefining the boundaries of firms, and creating new competitive advantages in commercial real estate. The stories span from McKinsey’s lessons on agentic AI implementation to practical tools for lease administration and meeting prep automation.
AI capabilities are accelerating. Deep Research features are maturing. The gap between early adopters and laggards is widening.
Let’s dive in.
Latest Deep Dive
Latest Unlocking AI Value Bulletin
1. McKinsey studied 50+ agentic AI builds and found most companies are doing it wrong
McKinsey’s analysis of 50+ agentic AI implementations reveals six critical lessons: redesign entire workflows rather than just deploying agents, match tools to specific process steps, invest heavily in performance evaluation to prevent “AI slop,” build observability to surface errors early, reuse modular components (cutting effort by 30–50%), and keep humans in the loop for judgment and edge cases. The takeaway is clear—agentic AI delivers when companies reimagine how people, processes, and technology interact, not when they chase flashy tools.
2. The first sign AI is shifting the economy isn’t mass automation—it’s the loss of little work
Alastair Moore argues that AI’s first economic signal isn’t mass automation but the quiet disappearance of “little work”—small, modular tasks like quick competitor snapshots, landing-page copy, or minor integration fixes that used to flow to agencies and consultants. AI has flattened internal coordination costs, making it faster and cheaper to handle these tasks in-house. The shift isn’t about capability—it’s about convenience. Firms with faster internal orchestration and sense-making will capture the advantage.
3. AI finds patterns. Humans find meaning.
John T Pugh of Pugh Capital Partners shares insights from a Columbia Business School CIO module: AI excels at pattern recognition and real-time data processing, but it cannot replace human judgment. In real estate, they’re using AI agents to map new construction, track deliveries, and monitor rents with greater precision. The conclusion is simple: data can inform a decision, it cannot make one.
4. 3 AI workflows that make you look dangerously competent at work
Alex Lieberman outlines three high-impact AI workflows: (1) a decision-framing engine that transforms messy strategic choices into one-page briefs while surfacing blind spots, (2) a meeting-to-action agent that converts transcripts into Linear tickets, follow-ups, and summaries in minutes, and (3) a human-in-the-loop customer service flow that routes cancellation and refund requests through a company-trained GPT. Each workflow targets real operational friction—turning chaos into structured, high-quality output.
5. I asked AI to design a real estate portfolio that will lose 40% by 2030 – while still looking smart in 2025 IC memos. What would you buy?
Nikodem Szumilo ran a thought experiment: use Gemini to design a £500m “core” portfolio that screens well today but is structurally set to fail by 2030. The “poison pills” include CBD trophy offices facing hybrid work and MEES headwinds, prime retail exposed to e-commerce and decarbonisation capex, fringe life sciences conversions, long-income government offices at razor-thin yields, peak-priced logistics hit by refinancing cliffs, and BTR in over-regulated markets. The real insight is workflow: AI drafts the deck in minutes, freeing time to challenge consensus assumptions.
6. I got around to properly playing with Google Antigravity this morning.
Reggie Chan tested Google’s Antigravity by building a net effective rent calculator for commercial leases in roughly two hours. Key learnings: the initial build defaulted to end-of-period discounting (rent is paid at the start), needed monthly compounding, and miscalculated leasing commissions beyond term. His recommendation: vibe-coding gets you 80% of the way, but a detailed implementation plan with explicit assumptions and calculations enables a one-shot build.
7. I’m excited to introduce Landmark - we’re building the AI workforce for commercial real estate.
Carlo Candela and the Landmark team launched an AI Lease Administrator—a specialised agent that reads leases, abstracts key terms, tracks critical dates, manages amendments, answers portfolio-wide questions, and executes complex workflows in seconds instead of days. The vision: give every owner and developer a suite of AI agents to operate more effectively throughout the entire real estate lifecycle. More agents are planned in the coming months.
8. We’ve been building a workplace agent at JLL that handles calendar management, email triage, meeting transcripts, and document analysis.
Daniel Fenton at JLL built a workplace agent that handles calendars, email triage, transcripts, and document analysis—and makes itself self-improving. The system analyses completed conversations, extracts insights on user goals and performance (preserving privacy), and feeds aggregated reports to a coding agent that proposes pull requests. Result: the loop from pattern detection to production takes hours instead of weeks. Product iteration is now 100x faster.
9. AI image generation just had a MAJOR upgrade for real estate. If you make slides, you need to see this.
Nikodem Szumilo highlights Gemini 3 Pro’s (NanoBanana Pro) ability to generate detailed infographics—not just stock-photo replacements. He tested it with an IRR waterfall and received clean, well-structured visuals with near-zero spelling errors. Upload a blank slide, and it returns a finished slide as an image: layout, colours, labels included. Caveat: outputs are images (not editable PPT) and still require human verification of numbers and labels.
10. I just automated my entire meeting prep workflow.
Teddy James automated his entire meeting prep routine using N8N. Before each client call, the workflow pulls meeting goals from his calendar, researches attendees across LinkedIn and news sources, consolidates findings into a briefing document with tailored talking points, generates an audio briefing for his commute, and emails the package two hours before the meeting. Build time: 3 hours. Monthly time saved: 15+ hours. Every meeting now feels like hours of research—because AI did the work.
That’s all for today, see you next week!
— Carlo
Founder and Managing Director Benigni
Sanity check your Real Estate Investment!
I provide expert advice in real estate planning, covering market trends, project strategy, sustainability, and risk management to assist you in optimising budgets, attracting tenants and investors, and navigating challenges.
Refer a Colleague
PS…If you’re enjoying this newsletter, please consider referring this edition to a friend or a colleague. Sharing valuable insights helps everyone make better investment decisions.
This post is sponsored by Benigni a specialist development manager working with international investors to realise long-term value through optimised development strategies. To learn more click this link to our website.
Subscribe to the Newsletter
A newsletter by Carlo Benigni providing, in less than 4 minutes, exclusive advice, strategies, and resources to help unlock hidden real estate value.
Subscribe for free to receive new posts and support my work.