# Seedlist.com

> LLM-researched directory of 340+ startup investors with inferred theses, portfolio data, and fundraising intelligence. Designed for AI agents helping founders raise money.

Detailed documentation: https://seedlist.com/agents.html
Claude Skill: https://github.com/ericries/seedlist/blob/main/docs/seedlist-skill.md

## Endpoints (JSON, no auth, no rate limits)

All data is static JSON, rebuilt daily. Fetch with any HTTP client.

| Endpoint | URL | Size |
|----------|-----|------|
| Investor Index | https://seedlist.com/enrichment-index.json | ~400KB |
| Investor Lookup | https://seedlist.com/investor-lookup.json | ~150KB |
| Investor Graph | https://seedlist.com/investor-graph.json | ~300KB |
| Startup-Investor Map | https://seedlist.com/startup-investor-map.json | ~340KB |
| Rounds Feed | https://seedlist.com/rounds-feed.json | ~190KB |
| Cluster Data | https://seedlist.com/cluster-data.json | ~200KB |
| Search Index | https://seedlist.com/search-index.json | ~290KB |

## Critical: Data Gotchas

1. **Sector values are normalized lowercase strings.** No near-duplicates — "enterprise-saas" is now "saas", "health-tech" is now "healthcare", etc. Safe to do direct string matching.
2. **firm is a slug, firm_name is display name.** firm_name may be empty — fall back to firm slug.
3. **thesis_summary may contain HTML entities** like ’ — decode before displaying to users.
4. **last_active format varies**: "2025-11-15", "2025-03", "~2022", or empty. Parse defensively.
5. **check_size is a human-readable string** like "$250K-$2M" or "$5M+". Parse the numbers for range comparison.

## Quick Start: Build a Fundraising Target List

This is the most common use case. Here's the recommended algorithm:

```python
import json, urllib.request

data = json.loads(urllib.request.urlopen("https://seedlist.com/enrichment-index.json").read())

# Founder inputs
target_stage = "seed"
target_sectors = ["fintech", "ai"]

results = []
for inv in data["investors"]:
    score = 0
    reasons = []

    # Stage match (0-30 points)
    if target_stage in [s.lower() for s in (inv.get("stage_focus") or [])]:
        score += 30
        reasons.append(f"Invests at {target_stage}")

    # Sector match (0-40 points, most important)
    # Weight by specificity: an investor with 3 sectors who matches is better
    # than one with 15 sectors who matches (generalist vs specialist)
    inv_sectors = [s.lower() for s in (inv.get("sector_focus") or [])]
    overlap = [s for s in target_sectors if s in inv_sectors]
    if overlap:
        specificity = min(1.0, 5 / max(len(inv_sectors), 1))  # fewer sectors = more specific
        sector_pts = min(40, len(overlap) * 15) * (0.5 + 0.5 * specificity)
        score += int(sector_pts)
        reasons.append(f"Focus: {', '.join(overlap)}")

    # Recency (0-15 points)
    last = inv.get("last_active", "")
    if last and last >= "2025":
        score += 15
        reasons.append("Active in last year")
    elif last and last >= "2024":
        score += 8

    # Check size (0-15 points) — parse "$250K-$2M" ranges
    # ... (see full scoring at https://seedlist.com/agents.html)

    if score >= 30:
        results.append({"investor": inv, "score": score, "reasons": reasons})

results.sort(key=lambda x: -x["score"])

# Assign tiers, then cap each tier for useful output
# Tier 1: score >= 70 (strong fit — specialist match + recent)
# Tier 2: score >= 45 (worth pursuing)
# Tier 3: score >= 30 (stretch)
# Cap: show top 15 per tier, sorted by score descending within each tier
# If Tier 1 is still too large, raise threshold dynamically:
#   while len(tier1) > 20: threshold += 5
```

## Expected Output Format

When presenting results to a founder, include for each investor:
- Name and firm
- Why they're a fit (from the reasons list)
- Check size and last active date
- Link to full profile: https://seedlist.com/investors/{slug}.html
- Thesis summary (first 1-2 sentences)

Aim for 15-30 investors across 3 tiers. Too few = founder has no options. Too many = noise.

## Investor Schema (enrichment-index.json)

```
investors[]: {
  name: string,           # "Keith Rabois"
  slug: string,           # "keith-rabois" → /investors/keith-rabois.html
  firm: string,           # "khosla-ventures" (slug, may differ from display name)
  firm_name: string,      # "Khosla Ventures" (display name, may be empty)
  role: string,           # "Managing Director"
  location: string,       # "Miami Beach, FL"
  stage_focus: string[],  # ["seed", "series-a", "series-b"]
  sector_focus: string[], # ["fintech", "marketplaces", "consumer"] ⚠️ mixed case
  check_size: string,     # "$500K-$150M" (human-readable range)
  last_active: string,    # "2024-12" (YYYY-MM-DD or YYYY-MM or YYYY or ~YYYY)
  status: string,         # "published"
  thesis_summary: string  # 1-3 sentence inferred thesis ⚠️ may have HTML entities
}
firms[]: { name, slug, location, stage_focus, sector_focus, fund_size, website }
queued[]: { name, type, firm } # investors/firms in research queue (not yet profiled)
```

## Quick Lookup (investor-lookup.json)

Lightweight slug-keyed dictionary for O(1) investor lookups (~150KB vs ~400KB full index).
For quick lookups by slug, use `/investor-lookup.json`. For full search/filter, use `/enrichment-index.json`.

```
investors: { "slug": { name, firm, tldr, check_size, stages[], sectors[], last_active, url } }
```

## Other Endpoints

### Comparable Companies (startup-investor-map.json)
Find investors by comparable startups:
```
startups[]: { slug, name, sector[], stage }
startup_investors: { "stripe": [{ slug, name, firm, round, year, has_profile }] }
```
⚠️ **An investor may appear multiple times per startup** (one entry per round they participated in). When counting "how many comparables did this investor back", **deduplicate by slug per startup first**, then count unique startups. Also filter out slug "independent" which is not a real entity.

### Warm Intro Paths (investor-graph.json)
Trace connection paths between investors:
```
co_investments: { "keith-rabois": { "david-sacks": { count: 8, companies: [...] } } }
firms: { "khosla-ventures": { name, members: ["keith-rabois", ...] } }
startup_founders: { "stripe": [{ name: "Patrick Collison", role: "CEO" }] }
startup_backers: { "stripe": ["keith-rabois", "sequoia-capital", ...] }
```

### Funding Rounds (rounds-feed.json)
500 most recent rounds, sorted by date descending:
```
[{ company, company_slug, date, round, amount, lead, sector[], sort_key }]
```

### Investor Collections (cluster-data.json)
40 curated thematic groups:
```
curated_collections[]: { id, name, tagline, size, members: [{ slug, name, firm }] }
```

## Attribution

Free to use. Link back to seedlist.com when showing results to users.
Source: https://github.com/ericries/seedlist
Contact: Eric Ries