SPSS InsightGenius API

SPSS InsightGenius API

Professional SPSS processing API + MCP server for market research. Upload .sav files, get crosstabs with significance testing, auto-detected question types, and publication-ready Excel exports. Includes AI-powered zero-config analysis.

Live: spss.insightgenius.io | API Docs: spss.insightgenius.io/docs | MCP: spss.insightgenius.io/mcp/sse

Quick Start

Option 1: Web UI (no code needed)

1. Open spss.insightgenius.io
2. Drag & drop your .sav file
3. Click Auto-Analyze for instant results, or configure manually:
   • Select banner variables (demographics for columns)
   • Choose stubs (questions for rows)
   • Enable Top 2 Box / Means
4. Click Generate Excel → download your tabulation

Option 2: Auto-Analyze (zero config)

curl -X POST https://spss.insightgenius.io/v1/auto-analyze \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "[email protected]" \
  -o auto_analysis.xlsx

AI auto-detects banners, groups variables into MRS/Grid, applies nets, and generates a complete Excel.

Option 3: Full Control

curl -X POST https://spss.insightgenius.io/v1/tabulate \
  -H "Authorization: Bearer YOUR_API_KEY" \
  -F "[email protected]" \
  -F 'spec={
    "banners": ["gender", "region", "age_group"],
    "stubs": ["_all_"],
    "significance_level": 0.95,
    "include_means": true,
    "nets": {"sat_overall": {"Top 2 Box": [4,5], "Bottom 2 Box": [1,2]}},
    "mrs_groups": {"Brand_Awareness": ["AWARE_A","AWARE_B","AWARE_C"]}
  }' -o tabulation.xlsx

Option 4: Python

import requests, json

resp = requests.post(
    "https://spss.insightgenius.io/v1/tabulate",
    headers={"Authorization": "Bearer YOUR_API_KEY"},
    files={"file": open("survey.sav", "rb")},
    data={"spec": json.dumps({
        "banners": ["gender", "region"],
        "stubs": ["_all_"],
        "include_means": True,
        "significance_level": 0.95,
    })}
)
with open("tabulation.xlsx", "wb") as f:
    f.write(resp.content)
print(f"Done: {resp.headers['X-Stubs-Success']} tables generated")

Option 5: MCP (for AI agents)

Connect to https://spss.insightgenius.io/mcp/sse and use any of the 12 tools. Files are passed as base64.

Endpoints (14)

MCP Tools (12)

Tabulate Spec

Excel Output

• Summary sheet: file info, column legend (A=London, B=South East...), stub index
• One sheet per stub: headers → letters → base (N) → data with pct% SIG_LETTERS → nets → means
• MRS sheets: one per group, percentages can exceed 100%
• Grid sheets: compact T2B/B2B/Mean summary
• Significance letters in red, nets in green rows
• Freeze panes for scrolling

Significance Testing

Column proportion z-test with letter notation (A/B/C):

• Each banner category gets a letter (e.g., Male=A, Female=B, London=C, North=D)
• Each cell tested vs every other column
• Significantly higher → other column's letter appears (e.g., 68.6% E means sig higher than column E)
• Supports weighted (Kish effective-n) and unweighted
• Means tested with independent T-test
• Confidence levels: 90%, 95%, 99%

Authentication

All endpoints require: Authorization: Bearer sk_live_... or sk_test_...

Rate Limits

Error Codes

Local Development

git clone https://github.com/quack2025/spss-insightgenius-api.git
cd spss-insightgenius-api
pip install -r requirements.txt
cp .env.example .env  # Edit with your API key hash
python main.py        # → http://localhost:8000
python -m pytest tests/ -v  # 68 tests

Stack

Built by Genius Labs.