How I Entered YC Twice with Two Distinct Ideas in Consecutive Years
Entrepreneurship often requires stepping into unfamiliar territory and building something meaningful from zero knowledge. Below is a reflective account of my path to YC twice, each time with a different idea launched in back-to-back years. I built these ventures while working as a software engineer at Uber’s Advanced Technologies Group, creating self-driving cars, and I started with no prior experience in healthcare or construction.
Intro: A two-company journey in YC
From 2018 to 2020, I launched two startups that were both accepted into Y Combinator. The first focused on computational drug discovery for Alzheimer’s disease (YC S18; the company is no longer active). The second targeted construction trade contractors with a CRM solution (YC S19). My approach was distinctive: I began with zero domain knowledge and learned through rigorous cycles of conversation, experimentation, and rapid launches.
The core of my method centers on talking to people. The conventional founder impulse asks, “What problem do I solve?” But in practice, the more productive question is, “What process can I follow to discover meaningful problems?” For founders without industry insider insights, identifying worthwhile problems is challenging. Consider unfamiliar domains like pest control or waste management and ask:
- What are the interesting problems in this space?
- Which problems will customers pay to solve?
- Why do people care about these problems?
- What is the root cause of these problems?
- What potential solutions exist, and why haven’t they been built?
- How could you launch a solution and gain initial traction?
The takeaway: without real insight, there is limited room for innovation. The path forward is to talk to a lot of people, to listen, and to let those conversations surface themes that can guide startup ideas.
Case in point: building a construction-tech company without construction experience
My cofounder and I conducted 20 in-depth conversations with construction professionals—owners, general contractors, project managers, estimators, and superintendents. We asked simple questions about daily workflows, frustrations, and the tools they used. Early responses highlighted recurring pain points such as:
- Disputes over subcontractor progress
- Time spent traveling to inspect sites due to quality concerns
- Redundant data entry across multiple spreadsheets
Over time, these recurring stories formed a pattern: costs and disputes, late payments, and onerous data management. These themes became the seed for our initial product ideas.
Key takeaway: the process of identifying promising startup ideas is not complex—talk to many people and let those conversations guide you. Conversations provide raw, unfiltered insight into how an industry operates, revealing day-to-day frustrations that hint at opportunities for innovation. Treat each conversation as a puzzle piece and your mission as assembling those pieces into a credible startup concept.
A crucial insight: a rigorous learning process beats relying on a single idea
Launching in industries where you have little experience presents a familiar challenge: how to attract people to share their wisdom when you don’t yet offer something valuable. Sourcing conversations is more of an art than a science, and it varies by industry, network, and circumstances. Despite that, the central discipline remains constant: build a repeatable process for learning and iterating toward impactful ideas.
From conversation to sandbox: the role of small, testable launches
Early on, learning from conversations can stagnate. To progress, you must launch a product, even if imperfect, to gain actual user feedback. Visionary founders can design products around a specific user empathy, but non-visionary founders often build for unfamiliar audiences. This means initial products may be misguided if they’re driven solely by conversation-derived intuition.
Key launch questions for early-stage products:
- Is the problem truly painful for the user, or does it feel painful to the user?
- Does the product address an industry-wide issue or a company-specific inefficiency?
- Is the solution effectively addressing the pain?
- Is the product easy to use?
- Can the product overcome entrenched processes and secure adoption?
A sandbox approach is essential for non-visionary founders. The goal is to launch multiple times, learn quickly, and move toward traction. I employed a concept I call a minimum launchable reason: a lightweight, credible justification to launch in a target industry that enables learning and iteration.
A practical playbook for sandboxes:
- Use conversations to identify an industry problem
- Seek industry partners to collaborate on solving that problem
- Outreach to partners proposing a project
The objective is to launch for learning, not to overpromise. For the two ventures I’m discussing, launches came through concrete, tangible partnerships:
- Drug-discovery project: volunteering to collaborate with a computational biology lab at Carnegie Mellon University
- Construction-tech project: partnering on a robotics initiative with a large construction company in Pittsburgh
While I’m omitting the specific outreach emails here, the essence is that a real-world project can unlock rich, practical insights—understanding data pipelines in biology or witnessing daily construction workflows and documentation practices firsthand.
From here to traction: a simple framework
- Identify problems by talking to experts and observing their routines
- Pinpoint a credible, painful problem
- Launch a product or project to learn rapidly
- Gather feedback and iterate
- Repeat until you achieve traction (e.g., LOIs, early revenue)
The reality of early-stage testing: ideas often fail, and that’s okay
It’s not unusual for early ideas to fail or for iterations to feel suboptimal. The critical factor is the speed of learning. My cofounder and I conducted three product launches across six months before settling on a current focus: a CRM for trade contractors. The timeline looked like this:
- First idea (autonomous robots for site scanning): four months to launch
- Second idea (construction analytics platform): one month to launch
- Third idea (CRM for trade contractors): one weekend to launch
We didn’t necessarily become smarter with each attempt. What improved was our ability to ship quickly and to devise creative ways to learn faster. Therefore, the guiding question shifted from “Is this a good idea?” to “What can I do to learn faster so I can position myself to discover a good idea?”
100% Founder Conviction: the defining milestone
Many people place emphasis on product-market fit, but I believe the pivotal milestone is 100% founder conviction—the moment when founders are unmistakably convinced that their idea will work. For our construction CRM, that moment arrived quickly after a slow start with earlier concepts. Within the first week of piloting with an electrical contractor, a demo, and a signed contract followed by multiple early customers. The early traction demonstrated that the problem was real and solvable, independent of the company’s size or appearance.
Of course, challenges remain:
- Securing continued traction before funds run out
- Convincing investors to provide seed capital
- Attracting talent to join a nascent team
The key difference is that these challenges feel manageable once founder conviction is established. The shift mirrors a transition from “Will this idea work?” to “This idea will work; we just need to assemble the proof points.”
Advice to my younger self
Looking back at the first seven years of entrepreneurship, the pace of progress often felt uncertain. Sleepless nights and moments of doubt are common. If I could offer guidance to my younger self, it would be to focus on two metrics:
- Rate of learning: Am I rapidly gaining knowledge about my space and idea?
- 100% founder conviction: Have I found that unwavering belief in the idea?
In retrospect, I would have spent less time consuming scaffolding content about fundraising or post-product-market-fit metrics, which are less relevant in the earliest stages. That said, I wrote this piece because I wished more founders would share how they discovered ideas and generated early traction. Origin stories have a powerful resonance and can inspire others with the potential to make a meaningful impact.
Final reflections
This account outlines a practical, process-driven path to discovering and validating startup ideas, even in domains where I started with no prior experience. The core lesson is simple: engage deeply with people, launch early and often, and cultivate 100% founder conviction as a north star. By treating conversations as the seeds of insight and constructing sandboxes for rapid learning, it’s possible to identify compelling opportunities and turn them into real, growing ventures. If you’re at the start of your own journey, consider how you can apply this framework to your industry and goals.
