$14M from OpenAI: The AI Agent That Lives in Excel and Transforms Real Estate Investing
Commercial real estate professionals waste 60% of their time on manual data entry and error-prone spreadsheets. Endex AI's Excel-integrated agent automates financial analysis with 90% error reduction - allowing investors to underwrite 10x more deals with the same team.
The Hidden Cost of Spreadsheet Errors in Real Estate
Commercial real estate professionals know the frustration all too well - you're reviewing a property's financials late at night, and something doesn't add up. The depreciation line items seem off, the COGS calculations for a nonprofit property make no sense, and suddenly you realize there are fundamental errors in the underlying spreadsheet.
This exact scenario inspired Endex AI founder Tarun's journey. While working with state boards on nonprofit funding allocations in high school, he discovered entire organizations fabricating their books through spreadsheet errors. "Some of these nonprofits just didn't exist," he recalls in the interview (12:45 timestamp).
Industry insight: A 2025 CRE industry survey found that 68% of underwriting models contain material errors, with the average team spending 22 hours per deal just on data validation.
How Endex AI Works: Your Excel Junior Analyst
The core premise of Endex is simple but revolutionary: an AI agent that lives inside Excel, learning your workflows and preferences like a junior analyst would. "Think about it as like a junior intern or analyst that lives in Excel that will start to learn your preferences and almost be an extension of you," explains Tarun (18:30 timestamp).
Unlike generic AI tools, Endex specializes in financial workflows with features like:
- PDF-to-model conversion for offering memorandums
- Automatic reconciliation of financial statements
- Tax code auditing against projections
- Scenario modeling beyond basic sensitivity tables
The $14M backing from OpenAI gives Endex access to frontier model capabilities before public release, allowing for financial-specific fine-tuning most tools can't match.
Transforming Commercial Real Estate Workflows
Commercial real estate professionals are notoriously resistant to changing their Excel-based workflows. Endex succeeds where other tools fail by meeting users where they already work rather than forcing migration to new platforms.
"Excel is not going anywhere," notes one podcast host (32:15 timestamp). "What makes Endex beautiful is that it integrates within your existing workflows." Key transformations include:
Before Endex: Analysts spend 60% of time on data entry and validation, with limited capacity for strategic analysis.
After Endex: Teams reallocate 80% of that time to higher-value activities like deal sourcing and relationship building.
Specific real estate use cases mentioned in the discussion include ADU feasibility analysis, lease vs buy comparisons, and automated rent comp analysis - all while working within existing institutional models.
10x Deal Underwriting Capacity
The most transformative potential lies in scaling underwriting capacity. "In the same amount of time it takes to underwrite one deal, I want Endex to help me underwrite 10 deals," explains one real estate investor (47:20 timestamp).
Early adopters report being able to:
- Process offering memorandums 8x faster via PDF extraction
- Run parallel scenario analyses on multiple properties simultaneously
- Automatically update models based on changing market conditions
One private equity firm cited in the discussion reduced their analyst-to-deal ratio from 1:3 to 1:15 within three months of implementation, without adding staff.
The 18-Month Roadmap to Human-Level Accuracy
While already impressive, Endex's capabilities are rapidly evolving. "I truly believe that in the next 18 months, Endex will match the precision of a $250K Harvard-educated analyst," states Tarun (52:40 timestamp).
The roadmap includes:
- Multi-worksheet template filling at 95%+ accuracy
- Proactive model updates based on external data changes
- Image analysis for property condition assessments
- Automated comps analysis across multiple data sources
These advancements will increasingly shift analysts from data processors to strategic decision-makers overseeing fleets of AI agents.
The Jarvis Vision: AI-Assisted Property Inspections
The most futuristic application discussed is what hosts called the "Jarvis vision" (1:02:30 timestamp) - conducting property inspections with AI assistance:
Future scenario: While walking through a property, you tell your AI agent "The walls need painting - add $5K to the rehab budget" or "The roof needs replacement." The agent updates the underwriting model in real-time, pulling comps and calculating ROI impact automatically.
Tarun predicts this level of integration within 30 months, combining:
- Voice interface for hands-free operation
- Image recognition for condition assessment
- Automatic data retrieval from trusted sources
- Continuous model updating based on new inputs
This would fundamentally change how investors evaluate and underwrite properties in the field.
Watch the Full Tutorial
See Endex AI in action during the full 30-minute discussion, including live demos of PDF extraction (starting at 22:10) and complex model population (38:45).
Frequently Asked Questions
Common questions about AI agents in Excel
Endex AI acts like a junior analyst living inside Excel that learns your workflows. It can extract data from PDFs, reconcile financials, run scenario analyses, and reduce spreadsheet errors by up to 90%.
The agent learns your preferences over time and can automate repetitive financial modeling tasks that normally require manual data entry and validation.
- Processes offering memorandums into populated models
- Automatically flags calculation discrepancies
- Runs parallel scenario testing
- Updates models based on changing market conditions
Unlike generic AI tools, Endex specializes in financial workflows within Excel. It understands complex multi-sheet models, financial terminology, and can process documents like offering memorandums.
The $14M OpenAI backing gives it access to frontier model capabilities before public release, allowing for financial-specific fine-tuning most tools can't match.
- Financial domain specialization
- Excel-native operation (no new interface)
- Access to cutting-edge model versions
- Continuous workflow learning
Endex can automate rent comp analysis, cash flow projections, sensitivity testing, lease vs buy analysis, and ADU feasibility studies.
It's particularly strong at extracting data from PDFs and populating complex underwriting templates with 95% accuracy, reducing the manual work typically required for these analyses.
- Automated rent roll processing
- Operating expense benchmarking
- Tax code compliance checking
- Capital expenditure planning
Currently matches junior analyst accuracy (85-90%) on structured financial tasks. The company projects human-level precision (95%+) within 18 months as models improve.
Key advantage is consistency - no fatigue or oversight errors that plague manual data entry. The AI maintains the same accuracy whether processing 1 deal or 100.
- Current accuracy: 85-90%
- Projected accuracy: 95%+ within 18 months
- Zero fatigue-related errors
- Continuous accuracy improvements
Not fully - it augments analysts by handling repetitive work. Teams report being able to underwrite 10x more deals with same staff.
The tool excels at data processing while humans focus on strategy, relationships and complex judgment calls that require human intuition and experience.
- Augments rather than replaces
- 10x productivity multiplier
- Humans focus on high-value work
- Better work-life balance for teams
Designed for Excel-fluent professionals with no coding needed. Users interact through natural language prompts in familiar Excel environment.
The average onboarding takes under 2 hours for basic functions, with full workflow mastery in 1-2 weeks of regular use. Most commands can be phrased as you would ask a junior analyst.
- No coding required
- Natural language interface
- 2-hour basic onboarding
- 1-2 weeks to full mastery
Connects to major CRE data providers like Costar, REIS, and proprietary databases through API integrations.
Future versions will automatically scrape public sources like Zillow and LoopNet. The system learns which data sources you trust for different analysis types and can blend multiple sources seamlessly.
- API connections to major providers
- Automated public data scraping
- Source reliability learning
- Data blending capabilities
GrowwStacks helps businesses implement AI automation tailored to their financial workflows. Our team designs solutions that integrate with your existing Excel models and processes.
We offer:
- Custom Excel automation solutions
- AI agent implementation
- Workflow optimization consulting
- Ongoing support and training
Book a free 30-minute consultation to discuss automating your financial analysis and underwriting processes.
Automate Your Real Estate Underwriting
Stop wasting time on manual data entry and error-prone spreadsheets. Let us help you implement AI automation that lets your team underwrite 10x more deals with the same staff.