theboylexis/momo-parse-api

Raw MoMo SMS → credit signals. Open API that turns MTN & Telecel transactions into structured financial intelligence for Ghana's lending ecosystem. # MomoParse — User-Owned Financial Intelligence from Mobile Money SMS Parse raw Mobile Money SMS messages from MTN and Telecel into structured financial data, categorize transactions with ML, and compute financial health indexes — all from user-owned SMS confirmations. **Live demo** — **API docs** ### Who this serves - **Thin-file borrowers** — Ghanaians with no formal bank history but years of MoMo SMS can now demonstrate income stability and repayment capacity to lenders. - **Licensed credit scoring entities & digital lenders** — programmatic access to standardized, categorized transaction data for risk models, without building parsing infrastructure from scratch. - **Users themselves** — transparent visibility into the financial profile telcos already compute internally but never share back. MomoParse is **infrastructure, not a lender.** Credit scoring is a regulated FinTech activity under Bank of Ghana licensing; MomoParse provides the structured data layer that licensed entities build on top of — keeping the user in control of their own SMS data. ```python import parser as p result = p.parse("0000015061132227 Confirmed. You have withdrawn GHS299.58 from A11205 - ACCRA TRADERS on 2025-09-10 at 13:51:07. Your Telecel Cash balance is GHS654.03.") print(result.telco) # "telecel" print(result.tx_type) # "cash_withdrawal" print(result.amount) # 299.58 print(result.balance) # 654.03 print(result.date) # "2025-09-10" print(result.confidence) # 0.9 ``` --- ## The Problem In Ghana's mobile money market (74.6M registered wallets, 23.9M active, GHS 1.00T in Q1 2025 alone — up 74% YoY, on track for ~GHS 4T annually), telcos score users with proprietary algorithms the user never sees. Credit scoring is an officially licensed FinTech activity (15% of 59 approved FinTech entities), yet users generate the data — telcos own the intelligence. MomoParse inverts this: it extracts **6 of the 8 standard alt-credit signals** from user-owned SMS d …