My YC Summer 2025 Rejection: The Complete Application

I applied to Y Combinator Summer 2025 with Edicek. I got rejected. And I'm sharing my full application so you can tell me what went wrong.

TL;DRYC Summer 2025 rejected my application for Edicek, a personal knowledge base that asks WHY you're saving content. I had almost zero user traction and I'm a solo founder who refuses to bring on a cofounder just to check a box. Here's the complete application - I want to know what to fix before I apply again.

Why I Applied

YC is the only accelerator that makes sense to me. Not primarily because of the money.

I've worked at VC-backed startups before. What I've seen is that the networking and industry knowledge are worth far more than the check itself. $500K for 7% equity is a good deal, but the real value is being connected to people who have skin in the game and actually want you to succeed because it's their money on the line.

The program structure is exactly what I need. Focused time, access to founders who've done it before, and a network that opens doors I can't open myself.

I was actually approached by London investors who found my open source project on GitHub. They said they invest in people who are consistent and build good products, and wanted to know if I was working on something they could invest in. Similar money range to YC. But talking to them, I realized the only thing they could offer was the money. That's not enough. That's why YC was the obvious choice.

Why I Think They Rejected Me

Almost zero user traction. This is probably the main reason. I have a working MVP, validated problem-solution fit, but only a handful of test users. The product isn't ready for real users yet - I know this, and they know this.

Solo founder. YC prefers teams. I get it statistically. But here's where I disagree with them: I'm not going to bring on a random cofounder just to check a box. Someone I don't know, who lives abroad, who isn't working with me daily, who doesn't share my vision - that's worse than being solo.

If the right person comes along, great. But I won't force it just because investors prefer teams. Even if YC told me they'd accept me if I found a cofounder, I wouldn't do it artificially.

Maybe the vision sounds like a cliché. "I have a clear vision for the product" - everyone says that. But I genuinely believe Edicek will find its audience over months and years. Whether that belief came through in the application, I don't know.

What I'd Do Differently

Next time, I'd get more outside perspective before submitting. I have contacts I've worked with on other projects - I should have run this by them first.

I also realized that distribution is the main thing. It's what YC definitely looks at. Building a great product is one thing, but getting it in front of people is everything. This is something I want to work on - figure out what works and what doesn't. The question is, if I figure out distribution myself, will I even need YC?

The Full Application

Here's everything I submitted. I'm not hiding anything - if there's something obviously wrong, I want to know.

Intro video

This is a one-minute video I recorded for the application. YC asks founders to introduce themselves and explain what they're building. I memorized the script because reading it was too obvious. It took me about 36 takes to get this version - English isn't my first language, so speaking naturally while remembering the script was harder than expected.

Demo video

A quick walkthrough of the product as it looked when I applied. This was the state of Edicek in August 2025 - rough UI, but working core features.

Application

COMPANY DESCRIPTION (50 characters max)
========================================
"If ChatGPT and Apple Notes had a baby"


WHAT IS YOUR COMPANY GOING TO MAKE?
========================================
The Problem: People save links, notes, images, videos, and documents but
never find them when needed. You bookmark articles, upload files, but when
you actually need that information, it's lost in digital clutter.

Our Solution: When users save content, we ask "Why are you saving this?"
This simple question changes everything - our AI creates focused summaries
(reading web pages, extracting video subtitles, analyzing documents) and
can bring relevant content back at the right moment based on your original
intent.

Why the "WHY" changes everything: Without context, AI results are
irrelevant. Save a product website because you like their design → AI
knows to suggest it for design inspiration, not product recommendations.
Without the "why," it's just guessing - and that's why every other
knowledge tool fails.

Example: Save a product website explaining "love their minimalist design
approach" → Later ask "Show me clean UI examples" → AI suggests that site
for design, not the actual product, because it remembers your original
purpose.

Current Product:
- AI search across all saved entries with intent understanding
- Plain text note creation with classification hints
- Bookmarklet for instant link saving with context
- Timeline view and collection organization
- AI chat powered by your personal knowledge base
- Credit system for transparent pay-per-use pricing

Next Steps (UX overhaul + content expansion):
- Professional UX redesign (current UI is rough MVP)
- Image/video uploads with AI extraction and analysis
- Web page backup for saved links
- OCR text extraction from images
- Video subtitle extraction and summarization
- Voice recording with automatic transcription
- Import tools (Notion, Obsidian, browser bookmarks, Instapaper)
- MCP server for Claude Desktop integration


WHO WRITES CODE, OR DOES OTHER TECHNICAL WORK ON YOUR PRODUCT?
========================================
Only me (founder) writes all the code and does all technical work. No
external developers or contractors have been involved.

I've built the entire technical stack including:
- GraphQL API with Pothos schema builder
- React frontend with TypeScript
- Prisma database layer with PostgreSQL
- Multi-layered authorization system
- AI integration with Anthropic Claude
- Bookmarklet and browser integration

The current MVP handles ~300 users and includes full-text search, AI chat
with knowledge context, collections management, and real-time content
saving.


ARE YOU LOOKING FOR A COFOUNDER?
========================================
I'm open to bringing on a cofounder, though not actively searching. The
product and technical execution are well-established, so if I were to add
a cofounder, it would be someone with complementary skills in business
development, marketing, or sales.

While I can handle these areas myself and have successfully built
product-market fit, having a dedicated business cofounder would allow me
to focus fully on product development and technical scaling, while they
drive user acquisition, partnership development, and market expansion.

The ideal cofounder would bring experience in B2C product marketing,
content creator/knowledge worker communities, or SaaS business
development.


WHERE DO YOU LIVE NOW?
========================================
Prague, Czech Republic / Prague, Czech Republic

Prague offers exceptional economic advantages for building a tech startup.
Development talent costs 3-4x less than Silicon Valley while maintaining
high quality - major companies like Avast, JetBrains, and Socialbakers
have proven Prague's tech ecosystem. Office space and operational costs
are significantly lower, extending our runway.

The city has a thriving startup scene with strong government support for
tech companies, EU market access, and a timezone that works well for both
US and European customers.


HOW FAR ALONG ARE YOU?
========================================
Current Status: Functional MVP with core features proven and few test
users.

What Works Today:
- Complete user authentication and workspace system
- Bookmarklet for instant content saving with classification hints
- Plain text notes creation
- AI-powered search across saved content
- AI chat that incorporates personal knowledge base context
- Collection organization and timeline views
- Credit system for transparent pay-per-use pricing

Validated: Few test users have confirmed core problem-solution fit. Users
successfully save content with intent explanations and retrieve it through
AI. Users don't mind writing the "why" - it's not a friction barrier when
they understand it improves retrieval quality.

UI Status: The current interface is quite weak - I'm fully aware of this
limitation. I intentionally prioritized building core functionality first
to validate whether the concept actually works. Now that the fundamental
features are proven functional, professional UI/UX is the immediate next
priority.

Market Research: Through 23 informal interviews, consistent patterns
emerged: most use Apple Notes and want additional functionality, but only
if it works invisibly in the background. Those who tried advanced tools
like Notion find them too complex and abandon them. Everyone uses ChatGPT
extensively but doesn't want to learn another tool.


HOW LONG HAVE YOU BEEN WORKING ON THIS?
========================================
I started actively building in May 2024 (3 months), working on weekends
and after work - not full-time yet. However, I've been thinking about this
product for about 1.5 years. My first notes are from February 2023,
documenting ideas and pain points around this exact problem.


ARE PEOPLE USING YOUR PRODUCT?
========================================
Technically yes, but realistically no. I've given access to a few people,
but I don't expect anyone to actively use it in its current state.

We're working with a graphic designer to design the interface from scratch
- everything currently visible is just rough backend scaffolding. We're
initially disabling some features and focusing purely on bookmarking,
search, and chat.

If I could work full-time on this, I estimate 3-4 months to have a
properly usable version that people would actively adopt and use daily.


WHY DID YOU PICK THIS IDEA?
========================================
Personal Pain Point:
I save around 3,000 bookmarks annually - videos, images, YouTube links,
articles, marketing strategies - but when I start working on projects, I
never actually utilize the content I've saved. Despite trying every
knowledge management tool available (Notion, Obsidian, Fabric, MyMind,
Raindrop, AirTable, Apple Notes), I still lose track of content I've
saved.

Domain Expertise:
I have 10+ years in software development. I understand both the technical
challenges of content management and the UX complexity of making powerful
tools feel simple.

Track Record:
Historically, I've excelled on projects that were in trouble or
stagnating. I'm good at identifying weak points that others miss, getting
into users' heads, and working closely with users to collect and apply
feedback.

The "WHY" Insight:
Through personal experience and user conversations, I discovered that
context (WHY you saved something) matters more than content analysis. AI
content extraction alone can't suggest the best solutions because it's
just guessing without knowing my original intent. This isn't obvious to
other tool builders, but it's the key to making saved content actually
useful months later.


WHO ARE YOUR COMPETITORS?
========================================
Direct Competitors: Fabric, MyMind, Raindrop, Supermemory, Lazy.so
Adjacent Competitors: Notion, Apple Notes, Readwise

What I understand that they don't:

These are all solid products solving real problems. But there's a gap they
don't address.

1. The Intent Context Gap: They focus on analyzing WHAT content contains,
but ignore WHY users saved it. Same article saved for "design inspiration"
vs "competitor research" needs completely different treatment.

2. The Market Fragmentation Opportunity: Total market is only ~$2-4M
monthly across ALL competitors. Raindrop has just 200K-500K users after 10
years. This isn't saturated - it's waiting for the right approach.

3. The Simplicity Paradox: They believe more features = more value, but
they're cutting off 80% of potential users who want Apple Notes simplicity
with invisible intelligence.

4. The Integration Imperative: They build walled gardens. Real adoption
happens when tools integrate into existing workflows (like MCP servers for
Claude).


HOW DO OR WILL YOU MAKE MONEY?
========================================
Pay-Per-Use AI Services

Current costs:
- AI search: ~$0.28 per operation
- Screenshot generation: ~$0.32-0.40 per screenshot
- Content storage, notes, bookmarks: Free
- AI chat: ~$0.28 per message

Revenue through transparent markup on actual AI costs. Users pay for what
they use - no subscriptions.

Target user economics:
- Casual users: $2-8/month
- Active users: $8-20/month
- Power users: Capped at ~$30/month

Market projections:
- Conservative: 10K users × $5/month = $600K ARR
- Target: 100K users × $8/month = $9.6M ARR
- Optimistic: 500K users × $8/month = $48M ARR


EQUITY STRUCTURE
========================================
I plan to form a Czech s.r.o. where I will initially hold 100% ownership
as the sole founder.

- Future co-founder: 15-30% depending on timing and contribution
- Employee equity pool: 10-15% reserved
- Investor equity: TBD based on funding rounds


WHAT CONVINCED YOU TO APPLY TO Y COMBINATOR?
========================================
I've known about YC for years, but despite having various ideas, I never
felt I had the right project to apply with. Friends and colleagues have
applied - some got in, some didn't - but I didn't want to apply just to
check a box.

This is the first project where I genuinely believe external funding is
necessary to build the product I envision. When I realized I needed
funding to execute properly, YC was the obvious choice.


HOW DID YOU HEAR ABOUT Y COMBINATOR?
========================================
Uhm, literally everywhere? :)

Tell Me What's Wrong

I'm applying again. I want to know what to fix.

Is the product unclear? Is the market opportunity not convincing? Is the lack of traction a dealbreaker no matter what? Is the solo founder thing actually that big of a deal?

If you've applied to YC, got in or got rejected, or if you just see something obviously broken in this application - I want to hear it.

Reach out at p.homoky@gmail.com. I'd rather hear harsh feedback now than make the same mistakes twice.

What Changed Since September (in 4 months)

• Over 1,000 changes to the codebase
• The web app looks completely different - new landing page, actual design instead of developer scaffolding
• Search is drastically faster and now free
• You can actually buy credits inside the app now
• We have proper documentation and a public API
• Mobile app is in progress, goal is January release, App Store in February
• New brand identity coming in January
• Preparing go-to-market strategy, probably going through TikTok creators
• Demo videos and tutorials in progress
• Started actively posting on @heyedicek
• Started building my personal brand on TikTok
• Traction is still zero

What's Next

Nothing changes. I keep going.

The rejection stings - no matter how much you tell yourself the odds are low, deep down you still hope you'll be the exception. But spending five days on that application forced me to think through things I'd been avoiding. Market size, strategy, competition, revenue model. Questions that were floating around finally had to be answered.

The application document is now my compass. The rejection was just an email.

---

If you made it this far - thank you. Seriously.

As a thanks, here's $50 in free credits for Edicek. After you sign up, click the three dots in the top right corner, go to Billing, and apply the code YC_REJECTION_50 in the Coupon section.