What It Actually Means to Run an AI-First Real Estate Company | Neal Bawa

Most real estate operators say they're using AI. Very few have built their entire company around it. In this episode, Neal Bawa, CEO of Grocapitus, breaks down the four-phase process his 20-person team used over the past 18 months to become what he calls an AI-first real estate company. From getting every employee certified on custom GPTs in phase one, to building proprietary web-hosted dashboards that pull live data from seven different property management systems in phase four, Neal is specific about what they built, what tools they used, how much it cost, and what it actually changed about how they operate. If you want to know what implementing AI at the company level looks like in practice, this is the episode.

About Neal Bawa

Neal Bawa is the CEO of Grocapitus and MultifamilyU, where he manages a $436 million AI-powered portfolio across 25 projects in 11 states. Known as the Mad Scientist of Multifamily, Neal has built one of the most data-driven operations in the real estate industry and has made more than 300 podcast appearances sharing that framework with investors. His free investor education platform at MultifamilyU has tens of thousands of subscribers and runs eight webinars per year at no cost.

What We Cover in This Episode

• What it actually means to declare your company AI-first and what changed internally at Grocapitus when they did
• How Neal tied AI adoption to compensation, bonuses, and continued employment — the carrot and stick system that drove company-wide adoption
• Phase one: getting every employee certified on ChatGPT and building 300 custom GPTs in four months
• Phase two: connecting Zoom, Slack, Asana, and Google Drive through native integrations and Zapier to automate meeting workflows
• Phase three: why they moved to Claude for rent comps, underwriting, and web scraping — and why ChatGPT couldn't do the job
• How Claude Code differs from Claude Chat and Claude Cowork, and why that distinction matters for automation
• Phase four: turning every employee into a programmer using GitHub Codespaces — no coding knowledge required
• How they built a proprietary MySQL dashboard that pulls live data from seven different property management systems simultaneously
• Why 99% of property managers are not compliant with the industry standard of three phone calls and three text messages to every lead — and how Neal can now see this in real time across his entire portfolio
• How AI affected headcount — no layoffs, but three to four open requisitions closed and employees working an average of one to one and a half hours less per day
• Phase five on the horizon: AI handling 100% of incoming calls at their properties
• Neal's plan to offer this dashboard-building capability to the top 50 property management companies in the US
• Why Haiku outperforms Sonnet for teams that keep running out of Claude tokens
• Neal's interest rate prediction through the end of 2026

Key Insight

Neal makes a claim that should stop most operators cold: 99% of leads at managed properties do not receive three phone calls and three text messages, the industry standard for lead follow-up. Most don't even get a second call. For years, this was invisible — property managers self-reported compliance, and nobody could verify it. Neal's phase four dashboard changes that. For the first time, he can walk into a Monday morning meeting and show every property manager exactly where they rank against each other, how long they take to respond to a lead, and how many of their leads are being processed correctly. He doesn't have to say a word. The data does it.

Why This Episode Matters

Neal isn't describing what AI might do for real estate someday. He's describing what his 20-person team built in the last 18 months, with zero consultants hired, on $25 per month Claude accounts. The playbook he lays out — phased adoption, compensation tied to AI competency,