Intro.
#Before You Start (3 min) — How to Bring In Your Business Plan
OpenSeed's current landing input accepts pasted business plan text and `.docx`, `.md`, `.markdown`, or `.txt` files. Direct PDF, PowerPoint, HWP, or HWPX uploads aren't supported, so you'll need to move the body text into a supported format.
| Original Format | How to Convert |
|---|
| Google Docs | File → Download → .txt or Markdown (.md) |
| Notion page | Export → Markdown (.md) — tables and images convert automatically |
| MS Word (.docx) | Upload the original .docx file as-is |
| PDF (not scanned) | Extract the text, then save as .txt (on Mac: Preview → select text → copy) |
| Only have handwritten notes or a slide deck | Write up the key content directly as .md (aim for 500+ characters) |
We recommend at least 1,500 characters of body text. Text under 500 characters doesn't give the 17 AI mentors enough signal to analyze, which lowers the accuracy of both the score and the comments.
02
#Steps 1–4 (5 min) — The 4 Founder-Context Inputs
The first four steps of the /start screen, beyond the body of your business plan, are context inputs that determine how the review criteria get calibrated. Each step takes an average of 30–60 seconds.
| Step | Input | How to Choose |
|---|
| 1. Founder type | Student · Prospective · Early-stage · Grant applicant (4 types) | If you haven't registered a business yet, choose 'Prospective'; if you're in years 1–3, choose 'Early-stage'; if you're seeking evaluation for a government grant program, choose 'Grant applicant' |
| 2. Industry | IT, biotech, fintech, content, social, etc. | Pick the single industry where your core revenue comes from |
| 3. Team | Solo · 2 people · 3+ people · incorporated or not | Even with a co-founder, list yourself as solo if you haven't signed a formal agreement yet |
| 4. Market | Domestic · domestic + overseas · overseas-focused | Base this on your current revenue/target market, not your future plans |
These four inputs adjust how the 17 AI mentors weight their evaluation. For example: a student founder + biotech + solo team + domestic market gets a deeper look at 'fit for Korea's Pre-Startup Package / Youth Startup Academy programs,' while an early-stage + fintech + 3-person team + overseas market gets a deeper look at 'fit for a Series A raise.'
03
#Step 5 (10 min) — Entering Your Business Plan Body
This is where you paste in your actual business plan text or drag in a .docx/.md/.markdown/.txt file. This is the step that takes the longest, and 80% of your result's accuracy is determined here.
- Minimum 1,500 characters, 3,000–5,000 characters recommended
- Keep any existing document structure as-is (headings, subheadings, bullet points, tables)
- If you're using a Pre-Startup Package or Early Startup Package template, the P-S-S-T structure works fine as-is
- For an IR deck, organize the text slide by slide, with the slide number attached
- Add text descriptions for images and charts (e.g., 'Graph 1: 2024 revenue ₩100M → 2025 revenue ₩300M')
File upload supports .docx/.md/.markdown/.txt. PDF, PowerPoint, HWP, and HWPX files aren't supported, so move the body text into a supported format and upload again.
04
#Step 6 → Analysis (5 min) — Payment and 15 Parallel Reviews
You request the analysis in Step 6 (available under the current payment terms shown on the analysis start screen). Once requested, analysis starts automatically, and 17 AI mentors work in parallel to do the following:
| Assigned To | Number of Mentors | What They Do |
|---|
| Core system | 7 | Standard evaluation across the 4 P-S-S-T axes, market viability, technical merit, team, and finances |
| Specialists | Up to 7 | Auto-matched by industry and stage (e.g., SaaS specialist, government-grant specialist, IR deck specialist) |
| IC Chair | 1 | Synthesizes all 17 mentors' comments into a final verdict |
The analysis itself typically finishes in 2–4 minutes. Even including payment processing time, you'll reach your report within 5 minutes. Even if you close the results screen, it's automatically saved to your account, so you can reopen it anytime from /history.
05
#Your Round 1 Report — Where to Start Reading
When the report opens, the first thing you see is the overall star rating and the verdict. But there's a specific reading order we recommend for first-time users.
- 1. **The one-line verdict** — an overall judgment like 'High chance of passing / Needs revision / At risk of rejection'
- 2. **The red flags section** — weaknesses that multiple mentors flagged in common (the reasons your score is lower)
- 3. **Category-by-category ratings** — which axis is weak (market viability, team, finances, etc.)
- 4. **Sentence-level rewrites** — concrete revision suggestions for weak sentences
- 5. **Comparison to successful applicants** — how you stack up against similar businesses that passed
90% of first-time users read in the order of overall score → detailed comments, but in practice, what actually helps most is starting with the red flags. You need to know 'why you lost points' before you can decide where to start revising.
06
#Moving to Round 2 — The Revise-and-Resubmit Cycle
If you fix 3–5 of the red flags from your Round 1 report in the body text and then run Round 2, the same 17 mentors evaluate it again against the same criteria. Because a consistency guarantee applies, the same sentence gets the same score.
| Round | Recommended Work | Expected Score Change |
|---|
| Round 1 | Submit as-is → check red flags | Establish your baseline |
| Round 2 | Fix 3–5 red flags, then resubmit | +5 to +10 points (with market and team improvements) |
| Round 3 | Tailor to a government-grant template / convert to an IR deck | +3 to +5 points (reaching the passing bar) |
From Round 2 onward, you can request a new analysis of the same business plan using the 'Re-review' button on the results screen. Since each round requests a fresh analysis, we recommend resubmitting only after making substantial revisions.
Summary.
#Common Sticking Points — First-Time User Troubleshooting
| Symptom | Cause | Fix |
|---|
| PDF, PowerPoint, HWP, or HWPX upload fails | .docx/.md/.markdown/.txt are currently supported | Extract the body text and convert it to a supported format |
| Analysis doesn't finish after 5+ minutes | Temporary load on the Anthropic API | Refresh at /history — results are saved automatically |
| Score comes back too low (50s) | Body text is too short (under 1,500 characters) | Expand the body text, then run Round 2 |
| Same business plan, different score | You may have selected a different founder type | The founder type in Step 1 determines the evaluation criteria |
| Analysis doesn't start after requesting it | Your draft may have expired | Start over from Step 1 — drafts auto-save for 24 hours |
CTA
Your first OpenSeed submission takes just 30 minutes. Fixing even 3–5 red flags after your Round 1 report noticeably raises your chances of passing. Get a consistent analysis from 17 AI mentors, under the current payment terms shown on the analysis start screen period.
You Can Get Your First Analysis Report in 30 Minutes
Convert your business plan to .md/.txt → complete the 6-step input flow → get a parallel analysis from 17 AI mentors.
Pre-submission business-plan review with clear service limitations
Start OpenSeed Review →