
Adam is the founder and CEO of AKTUS AI, a Silicon Valley company building custom AI systems for finance and real estate — and a firm Trion has been working with directly. He has a PhD in computer science, spent 15 years in AI and machine learning, and was product lead for data and ML at Twitter before co-founding Flowcode.
In this episode, David and Adam get into why out-of-the-box AI keeps failing on real estate work, what actually separates a prototype from a production system, and why Adam thinks the org chart itself is the thing AI ends up rewriting.
What we get into
Why generic AI can't do real estate. The models were trained on public data. Real estate runs on proprietary data, domain judgment, and a secret sauce that was never on the internet — and it's multidisciplinary on top of that: finance, legal, architectural, relationship-driven.
Vibe coding: 100% and 0%. Adam's line is blunt. Vibe coding is 100% the right tool for prototyping and requirement gathering — it replaced weeks of whiteboarding and Figma mockups. For production code, it's 0%. None of it survives. He explains why, and it comes down to edge cases and maintainability.
The maintenance trap. David's own experience: building something that works, then spending half your day keeping it alive instead of doing the job. The productivity gain never shows up.
Digital teammates, not software. You're not buying an application. You're onboarding something you train, give feedback to, and measure — the same way you'd manage a first-year analyst.
How you actually measure success. Booking a meeting is not the same as booking the right meeting. Adam walks through how AKTUS defines KPIs with a client during onboarding, and why measurement has to be designed in from day one rather than bolted on.
Why AI hallucinates, explained simply. Context windows, the myth of the million-token window (only about 10% of it is effective), and why AI forgets the middle of a document the same way you forget the middle of a novel.
Context engineering — and hiding data on purpose. Sometimes the right move is keeping data away from the model. Adam explains the system AKTUS built to ingest thousands of leases, zoning docs, and a 60-tab underwriting model without the whole thing falling apart.
Five years to autopilot. Today AI recommends and a senior analyst edits. Adam's estimate for when AI makes the call itself — and why the answer depends entirely on how much high-quality decision data you've logged.
Can AI invent something better than Excel? AKTUS is training AI to operate inside Excel. The next frontier is AI inventing the next abstraction layer entirely.
Inside the Trion project. David breaks down the lead gen platform they're building with AKTUS — power-ranking properties against portfolio performance, drafting genuinely customized outreach, and killing spray and pray. Plus why AI drafts the email but doesn't send it.
Overhyped and underhyped. Adam's answer: like the internet, AI is overhyped in the short term and underhyped in the long term. Specifically overhyped right now — drag-and-drop, one-shot deal screening.
Chapters:
00:17 Introduction to AKTUS AI
04:35 "Out-of-the-Box" AI Flaws in CRE
07:19 The Role of Custom AI in Real Estate
10:52 Where Vibe Coding Ends and Real Engineering Begins
18:01 Iterative Learning and Decision-Making in AI
18:31 Measuring Success in AI Implementations
23:27 Future of AI in Financial Modeling
25:04 Five Years Until AI Edits the Model Itself
30:18 Real-Use Cases of AKTUS AI: 1,000s of Docs, 60-Tab Models, & More.
35:42 Context Management and AI Limitations
39:37 Digital Teammates: Redefining Roles in Real Estate
44:49 Overhyped vs. Underhyped AI in Real Estate
46:51 The Evolution of Organizational Structures with AI
www.dealflowfriday.com
IG: @dealflowfriday
X:@dealflowfriday
More episodes from "Deal Flow Friday"



Don't miss an episode of “Deal Flow Friday” and subscribe to it in the GetPodcast app.








