Voice Memos (/api/v1/voice-memo)

Convert raw voice transcripts into polished, structured documents and professional PDF reports.

All endpoints require authentication via X-API-Key and consume credits (VOICE_MEMO).

---

1. Polish Voice Memo

Polishes a raw transcript into a well-formatted document with sections.

• Method: POST
• Endpoint: /api/v1/voice-memo/polish

Request Body

Parameter	Type	Description
transcript	String	Raw voice transcript (min 3 words)
doc_type	String	Document type: credit_memo, meeting_notes, due_diligence, risk_assessment

{
  "transcript": "Met with the borrower today. They are requesting a $500k loan for their manufacturing business. Revenue has grown 20% year over year. Cash flow looks strong...",
  "doc_type": "credit_memo"
}

Response

{
  "polished": "# Credit Memo\n\n**Borrower:** Manufacturing Co.\n**Loan Amount:** $500,000\n...",
  "title": "Manufacturing Co. - Credit Memo",
  "subtitle": "Loan Assessment Summary",
  "summary": "The borrower operates a profitable manufacturing business with consistent revenue growth...",
  "sections": ["Executive Summary", "Financial Analysis", "Risk Assessment", "Recommendation"],
  "word_count_raw": 120,
  "word_count_polished": 350
}

---

2. Generate Voice Memo PDF Report

Polish the transcript and generate a professional PDF report.

• Method: POST
• Endpoint: /api/v1/voice-memo/report

Request Body

{
  "transcript": "Raw transcript text here...",
  "doc_type": "credit_memo"
}

Response

• Content-Type: application/pdf
• Returns: Binary PDF stream of the polished voice memo report.