Building PropFlow: The Journey from Idea to MVP

Bringing PropFlow to life has been a challenge. A challenge that I put hours of work into as my co-founder Alex Taylor and I failed constantly—and tried again.

Creating software seems easy at first until it's time to get into the little details. Details like connecting databases, making real API calls to third-party services, and visualizing that data in a user-friendly way. Not to mention the difficulty of building secure authentication systems that protect user data while remaining valid across sessions.

The Pivot: Taking Control

Through this process, I learned so much about our tech stack. We started small with GitHub and Netlify deployments. But at one point, we knew we had to step up. We wanted more control.

That's when we made the change to Google Cloud and Firebase.

The challenge of tearing everything down and building it back up was immense. It was a major pivot; the system before "worked," but not well enough for the scale and reliability we needed.

The Deployment Struggle

During that transition, we lost our automatic deployment pipelines. Everything became much more hands-on. The amount of time our deployments failed was unfathomable.

To fix these issues, I had to create workflows and AI agents to sync our entire tech stack together. When you finally get past days of debugging and issues, it’s the most rewarding feeling in the world.

The Breakthrough: Custom AI Agents

Building with AI seems easy until you hit roadblocks. That’s why I felt it was so important to build specific workflows and sub-agents.

One breakthrough I had was taking a step back and realizing the real problem: My IDE did not have enough context.

Being new to this level of software engineering, I didn't always have the knowledge to even understand the issues I was running into. So, I turned to NoteLM (Google's research tool). I dove deep into:

• Best prompting practices
• Google Cloud and Firebase ecosystem details
• Best practices for scalable software architecture

NoteLM gave me precise instructions and clear insights like an expert. But I went a step further.

Now, simple descriptions of issues turned into 10-fold efficiency gains. Instead of guessing why authentication wasn't working, I had an agent telling me exactly "Google Cloud API database isn't synced with passkeys" (hypothetically).

The point is: I now had control. Instead of guesswork, I was learning best practices from an expert I built myself.

What is PropFlow?

Project Title: PropFlow - AI-Powered Market Intelligence for Property Management

MVP Title: Financial and Communication Optimization & Tenant Churn Prevention System

The Problem

Independent landlords and small property management companies face three critical challenges:

1. Manual Market Research is Time-Consuming
   Landlords spend 15-20 hours per month researching rent prices. Pricing based on "gut feeling" leaves $100-300/month on the table per property.
2. Tenant Turnover is Costly
   Each turnover costs $4,000-6,000 in vacancy, cleaning, and advertising. There is currently no early warning system for tenant dissatisfaction.
3. Lack of Actionable Intelligence
   Existing platforms focus on operations, not analytics. Investment decisions are made without comprehensive data.

The PropFlow Solution

PropFlow is the first AI-native platform that makes landlords money through intelligent automation, offering proactive AI automation, predictive analytics, and real-time market intelligence.

Team

• AT
  Alexander Taylor

  Primary Contact • ajt6597@psu.edu
• JT
  Joel Torres

  Co-Founder