Get Started in 5 Minutes

From signup to your first AI-powered deal evaluation — all from your terminal.

1. Sign up and get your API key

Create an account at creditoptix.com. Your API key is auto-generated — retrieve it from the onboarding status endpoint.

curl https://creditoptix.com/api/v1/onboard/status \
  -H "Authorization: Bearer co_live_YOUR_KEY"

# Response includes your API key on first call:
# { "api_key": "co_live_...", "step": "profile", ... }

2. Set your lender profile

Tell us what type of lending you do. This configures your AI agent's domain expertise.

curl -X POST https://creditoptix.com/api/v1/onboard/profile \
  -H "Authorization: Bearer co_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "lender_type": "mca",
    "name": "My Lending Co"
  }'

3. Upload training data (or use sample)

Upload your historical deals as CSV, or use our sample dataset to get started immediately.

# Option A: Use sample data (fastest)
curl -X POST https://creditoptix.com/api/v1/onboard/sample-data \
  -H "Authorization: Bearer co_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{}'

# Option B: Upload your own CSV
curl -X POST https://creditoptix.com/api/v1/onboard/upload \
  -H "Authorization: Bearer co_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "file_name": "deals.csv",
    "data": "'$(base64 deals.csv)'"
  }'

4. Add underwriting guidelines (optional)

Tell the AI about your specific lending criteria, rate sheets, or compliance rules. This shapes how the agent evaluates deals.

curl -X POST https://creditoptix.com/api/v1/onboard/doctrine \
  -H "Authorization: Bearer co_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "guidelines": "Minimum 2 years in business. No advance over 1.5x monthly revenue. Decline restaurants under $50k monthly."
  }'

5. Train your AI agent

Genesis analyzes your historical deals and builds a custom underwriting model. Takes 30-60 seconds.

# Trigger training
curl -X POST https://creditoptix.com/api/v1/onboard/genesis \
  -H "Authorization: Bearer co_live_YOUR_KEY"

# Poll until complete
while true; do
  STATUS=$(curl -s https://creditoptix.com/api/v1/onboard/status \
    -H "Authorization: Bearer co_live_YOUR_KEY" | jq -r '.agent_status')
  echo "Status: $STATUS"
  [ "$STATUS" = "active" ] && break
  sleep 5
done

# Activate the agent
curl -X POST https://creditoptix.com/api/v1/onboard/activate \
  -H "Authorization: Bearer co_live_YOUR_KEY"

6. Submit your first deal

Send a deal for AI evaluation. Get a recommendation with full reasoning in ~30 seconds.

curl -X POST https://creditoptix.com/api/v1/deals \
  -H "Authorization: Bearer co_live_YOUR_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "application_data": {
      "business_name": "Acme Corp",
      "amount_requested": 50000,
      "monthly_revenue": 85000,
      "credit_score": 680,
      "years_in_business": 4,
      "industry": "Retail",
      "state": "NY"
    }
  }'

7. Read the AI recommendation

Poll until the evaluation is ready, then read the recommendation, confidence score, and reasoning.

DEAL_ID="your-deal-id"
curl https://creditoptix.com/api/v1/deals/$DEAL_ID \
  -H "Authorization: Bearer co_live_YOUR_KEY" | jq '{
    recommendation: .deal.ai_recommendation,
    confidence: .deal.ai_confidence_score,
    reasoning: .deal.ai_reasoning
  }'

What's next?

Full API Reference — all 20 endpoints, schemas, and examples
npm install @creditoptix/sdk — TypeScript SDK with full onboarding, deals, documents, and agent methods
pip install creditoptix — Python SDK with the same complete API surface
Attach documents (bank statements, credit reports) via POST /deals/:id/documents for deeper AI analysis
Monitor your agent's accuracy and manage versions via /agent/status and /agent/history
Record deal outcomes via PATCH /deals/:id/status to improve agent accuracy over time