Thesis

My journey began with a personal observation of the massive “technological gap” within my father’s family business, which I viewed through the lens of my professional background as a strategy consultant. I saw a painful reality: while large enterprises have the capital and workforce to invest in custom ERP and CRM integrations, smaller businesses are trapped. They are forced to choose between rigid, expensive software they can’t afford or fragmented, manual workflows that fundamentally stall their growth.

I envisioned a “Wix-like” solution for business management: a modular, drag-and-drop SaaS platform that would empower non-technical owners to build their own internal infrastructure. The “magic” lies in a Natural Language Chat interface. Instead of hiring a consultant for months to configure custom software, a business owner could simply “talk” to the system to generate functional tools for invoicing, inventory, or logistics. Inspired by LLMs like Base44 or Lovable, I wanted to build a system that translated natural language into automated, functional business workflows.

The core of my thesis was to solve the “Rigidity Problem”. Typically, when an SMB adopts a new tool, they have to warp their successful, existing workflows to fit the software’s limitations. My goal was to flip that paradigm. By using AI to translate a business’s unique requirements into technical reality, I wanted the software to finally adapt to the business, rather than the other way around.

Validation

I approached validation with a rigorous, data-heavy mindset, knowing a “good idea” on paper is a far cry from a viable business in the trenches. I participated in an accelerator that proved invaluable for building a professional network and adding structure to my process. By leveraging my personal network, I conducted over 30 in-depth interviews with SMB owners and managers across various sectors to see if my hypothesis held water.

From the beginning, I focused on daily workflows to uncover the most painful and uncomfortable frictions, not just whether the idea sounded interesting. Eventually, I asked questions like, “Would you allocate a budget for this?” and “What would it take to replace your current system?” This revealed the real constraints: switching costs, integration pain, and ownership of spend. To map the landscape, I distributed a market survey to pinpoint exactly where digital infrastructures were failing. My goal was to reach dozens of businesses ranging from three-person shops to 300-employees; ultimately, 50 replied.

I also sought high-level expert feedback by pitching the idea to veteran industry leaders, including former CEOs of Israeli-based global tech giants. While they validated that the market gap was real, they offered a sobering warning about the massive early-stage risks of building a “backbone” product. To test the technical efficiency of my approach, I built a functional prototype using existing tools within my father’s business. In just five days, the system was managing complex tasks that previously required expensive manual labor, proving the tech could deliver. Eventually, my product was running in three SMBs.

I met with several angels and one VC, whose feedback was that it was simply too early; they needed more data and a working MVP. Throughout this journey, I stayed rooted in the local tech ecosystem, using accelerators, and a strategic mentoring hub to refine my vision.

Why it didn’t progress

1. Rapid Technological Obsolescence: As General AI like ChatGPT and specialized agents advanced at breakneck speed, I realized my unique “moat” was being commoditized. Large platforms were already beginning to integrate the very API-driven AI features I had planned.

2. The “All-or-Nothing” Product Trap: A business management system is a company’s “backbone”. I realized I couldn’t release a “partial” MVP; if the system wasn’t 100% robust and polished, businesses wouldn’t risk their operations on it. There was no room for “agile” mistakes.

3. The CTO Dilemma: Finding a partner was incredibly difficult. I interviewed several candidates I met through LinkedIn, but encountered a recurring “arrogance” in the tech market. The work pace and communication between my final CTO hire and me didn’t match what I was aiming for, which led to frustration and slower progress.

4. Fundraising Trap: While VCs demanded a polished MVP and active users, building a robust “backbone” system for SMBs required significant upfront capital that I couldn’t secure without that very MVP already in place.

Insights

1. Don’t Fall in Love with the Solution: You must remain objective. It’s easy to remain “blinded” by our own ideas. I chose to stop when the data showed the window of opportunity was closing.

2. The SMB Knowledge Gap is a Goldmine: The real value isn’t just the software; it’s the consultative layer. SMBs don’t just need a tool; they need a tool that teaches them to run efficiently, and within the SMB domain, inefficiencies can be found everywhere.

3. Founding Chemistry is Non-Negotiable: A part-time or “mercenary” CTO cannot build a “backbone” product. Technical leadership must be as obsessed with the domain and timeline as the business lead.

4. Watch the Tailwinds: If you build on top of rapidly evolving technology like LLMs, you risk being “eaten” by the platform providers before you even launch.

---