Document intelligence for Kenya lenders
Turn financial documents into lending decisions in minutes
Upload M-PESA statements, bank statements, and financial documents. Textrion extracts, verifies, scores, and prepares committee-ready credit reports automatically.
Upload statement
Activejane_mpesa_jan-mar.pdf
M-PESA Statement
3 months · PDF
OCR extracts & verifies
142 transactions parsed
Credit signals extracted
Ready for review
Credit report ready
Jane M.
Integrity score
82/100
Income
KES 124,000
Monthly expenses
KES 61,000
Debt ratio
23%
Try with a sample document · No signup · Jane M. demo case
PDF & batch upload
Drop M-PESA and bank PDFs from the branch. Hybrid text + OCR handles scans and phone photos.
WhatsApp intake
Borrowers forward statements on WhatsApp. Textrion extracts and links each file to the borrower case.
Email forwarding
Dedicated inbound addresses per workspace. Attachments become structured extractions automatically.
API & webhooks
Push completed extractions to your LOS, spreadsheet, or automation stack with one integration.
How it works
Four steps from document to lending decision
No spreadsheets, no re-keying. Upload a statement and get a committee-ready credit file in minutes.
Step 1 of 4 · live now
Upload
Drop M-PESA PDFs, bank statements, or forward from WhatsApp and email.
Live product proof
See a real borrower case, no account required
M-PESA + KCB statements, integrity grades, combined credit review, and committee memo: exactly what your loan officers get after extraction.
Who is this for?
For lenders — and for anyone who wants clarity on their M-PESA
Loan teams use Textrion for borrower files. Individuals use personal analysis on their own statements.
Individuals
Upload your own M-PESA statement and see spending, cash flow, and wellbeing tips.
Banks
Salaried borrowers with bank + mobile money statements.
Microfinance
High-volume applications with M-PESA-heavy files.
SACCOs
Member lending with branch intake and central review.
SME lenders
Business owners with mixed financial documents.
Credit analysts
Officers who need structured data, not PDF hunting.
For individuals
Just want to understand your own M-PESA spending?
Register as an individual — no team, no work email. Upload your Safaricom statement and browse almost everything Personal analysis surfaces: spend, people & paybills, standouts, and commitments.
Spending by category
See where your M-PESA money actually went — groceries, transport, paybills, Fuliza, and more.
People, paybills & money in
Top paybills, who you send to most, and where money usually comes from — with how often.
Standouts & commitments
What stands out, monthly bills, fixed vs flexible spend, and steps toward financial freedom.
Includes 50 free page credits and 15 days free Personal analysis & Credit review.
Personal analysis
Cash flow · last 30 days
Inflows
KES 48.2k
Outflows
KES 44.7k
Savings rate
7.3%
Where the money went
- Food & groceriesKES 12.4k · 28%
- TransportKES 8.1k · 18%
- Send moneyKES 9.8k · 22%
- Paybill & tillKES 6.8k · 15%
- Airtime & dataKES 4.1k · 9%
- OtherKES 3.6k · 8%
Tip: Cut weekend till spend by ~KES 1,500 and your savings rate doubles this month.
Extraction in action
See the transformation: statement → structured credit data
Left is what your officer receives. Right is what Textrion produces in under two minutes — OCR plus deterministic parsing, without sending documents to generative AI.
M-PESA STATEMENT
Jane Mwangi · Jan–Mar 2024
Opening balance: KES 12,450.00
| Date | Details | Amount |
|---|---|---|
| 15 Jan | Funds received from ACME LTD payroll | +105,000 |
| 05 Jan | Pay Bill KPLC | -15,000 |
| 12 Jan | Send Money to JANE DOE | -8,500 |
| 18 Jan | Fuliza repayment | -2,100 |
… 138 more rows in PDF
{
"document_type": "mpesa_statement",
"borrower_name": "Jane Mwangi",
"period": "Jan–Mar 2024",
"transactions": [
{ "date": "2024-01-15", "type": "credit",
"amount": 105000, "party": "ACME LTD" },
{ "date": "2024-01-05", "type": "debit",
"amount": 15000, "party": "KPLC" }
],
"integrity": { "grade": "good", "score": 0.94 },
"flows": {
"avg_monthly_in": 124000,
"avg_monthly_out": 61000
}
}Inside the product
What you get after upload
Dashboard, extraction, credit review, and API delivery. No imagination required.
Runs (30d)
248
Pages
1,420
Success
97%
OCR pages
86
Runs, pages, and success rate at a glance
Drop M-PESA or bank PDFs
jane_mpesa.pdf · kcb_stmt.pdf
Upload & extract
Drag-and-drop PDFs from the branch
142 transactions · Integrity: Good
Live extraction with integrity grading
Avg monthly in KES 124,000 · Band KES 18–22k
Stable payroll inflows with predictable obligations…
Repayment signals and recommendation
| Date | Description | Cr |
|---|---|---|
| Jan 15 | ACME payroll | 105k |
| Jan 05 | KPLC | -15k |
Searchable credit and debit rows
{
"integrity_grade": "good",
"avg_monthly_in": 124000,
"comfortable_installment": [18000, 22000]
}Structured JSON for your LOS
Analyze your first statement free · No credit card required
Trust & security
Financial teams ask hard questions. We answer them upfront.
Encrypted uploads
TLS in transit for every file and API call.
Retention you control
Team workspaces with configurable document retention.
Privacy-first design
Role-based access. Your underwriting rules stay in control.
Secure processing
Encryption at rest on your hosting provider.
Supported documents
Every document your credit team already receives
Dedicated parsers where lending depends on them. Generic extraction with confidence warnings everywhere else.
M-PESA statement
Safaricom PDFs and phone photos
Bank statement
Co-op, KCB, Equity, and 15+ Kenya banks
Payslip
Salary verification with field confidence
Financial statements
Business accounts and supporting docs
ID documents
National ID and identity extraction
Business registration
CR12, certificates, and filings
Use cases
Personal clarity or lender-grade credit files
Individuals decode their own M-PESA. Credit teams turn statements into integrity grades and committee memos.
Personal spending analysis
- 1Register as individual
- 2Upload M-PESA PDF
- 3See categories & cash flow
- 4Act on tips
Loan underwriting
- 1Upload statements
- 2Income analysis
- 3Risk signals
- 4Committee memo
Fraud & integrity checks
- 1Upload statement
- 2Balance reconciliation
- 3Anomaly detection
- 4Risk alerts
Co-op, KCB, Equity & more
From PDF to committee memo
Balance, gaps & identity checks
M-PESA + bank + combined review
Committee memo
Your policy. Our narrative.
Textrion doesn't replace your credit scoring. It turns verified statements into the committee-ready story your officers were writing by hand.
Statement integrity
Balance reconciliation, date gaps, and identity checks, graded before your officer signs off.
- →Not a bureau score. Your underwriting rules stay in control
- →Proceed / review / caution recommendation on every combined case
- →Exported in credit review PDF and borrower dossier
WhatsApp intake
Borrower forwards a statement. Officer gets a case.
Field agents and branches already work on WhatsApp. Textrion meets borrowers where they are, then links every extraction to the borrower file.
Good morning, attaching my M-PESA statement for the loan application.
Received. Processing now. You'll get a case link shortly.
✓ Extracted · Integrity: Good · Linked to case Jane M., Jan–Mar 2024
Case ready for review. Committee memo drafted. Open in Textrion → Open demo
Document coverage
Kenya statements your team actually receives
Dedicated parsers where it matters for lending. Generic extraction with confidence warnings everywhere else; we flag weak fields instead of pretending.
| Document | Formats | Lending workflow | Notes |
|---|---|---|---|
M-PESA statement | Text PDF · Scanned PDF (OCR) · Phone photo | Core | Personal and merchant till (shortcode) statements. Regression-locked personal path; till statements get business cash-flow analysis. |
Kenya bank statements | Core | Dedicated parsers per bank profile below; additional banks use generic table extraction. | |
Invoices | Platform | Line items, totals, and vendor fields for cloud and services invoices. | |
Receipts | PDF · Image | Platform | Merchant, amount, and date extraction. |
Bank profiles with dedicated parsers (9)
Other banks → generic table extractionUnsupported or low-confidence extractions are flagged for officer review. Textrion does not auto-approve limits.
ROI estimate
How much officer time comes back?
Adjust for your branch volume. Most pilots see the biggest win on statement review and memo drafting, not extraction alone.
Estimated time saved
Illustrative only, based on your inputs. Actual savings depend on document quality, case complexity, and team workflow.
Outcomes lenders care about
Every feature answers a business question
OCR and rule-based analysis in your workspace — not third-party AI. Signals your credit committee can defend.
Verify statement authenticity
Balance reconciliation, date gaps, and identity checks graded before approval.
Detect salary patterns
Payroll inflows, informal income, and volatility surfaced for credit narrative.
Calculate affordability
Comfortable installment bands from verified paid-in and obligation data.
Identify existing loans
Recurring debits, Fuliza, and paybill obligations tagged automatically.
Generate committee-ready reports
Auto memos with proceed / review recommendations your team can defend.
Export structured data
CSV, XLSX, JSON, credit review PDF, and borrower dossier per case.
API integration
Webhooks and REST endpoints when extraction completes in your LOS.
Bulk document processing
Batch uploads, WhatsApp intake, and email forwarding at branch scale.
FAQ
Questions your buying committee will ask
Straight answers for credit, operations, and IT, before you run a pilot.
Still evaluating? Open the live sample case or book a walkthrough.
Pricing
Start free. Buy credits when you need more.
50 free page credits on signup, plus 15 days free Credit review and Personal analysis. Prepaid packs never expire — volume discounts from KES 5 to KES 3 per page. Enterprise gets custom rates. Prices in Kenyan shillings.
50 free pages on signup · never expires
- ✓50 extraction pages included
- ✓15 days free Credit review & Personal analysis
- ✓No card required to start
- ✓Buy more credits anytime
≈ KES 5 per page · never expires
- ✓50 extraction pages
- ✓Use anytime — no monthly reset
- ✓Pay with Paystack (M-Pesa, card, bank)
≈ KES 5 per page · never expires
- ✓100 extraction pages
- ✓Use anytime — no monthly reset
- ✓Pay with Paystack (M-Pesa, card, bank)
≈ KES 4 per page · never expires
- ✓500 extraction pages
- ✓Use anytime — no monthly reset
- ✓Pay with Paystack (M-Pesa, card, bank)
≈ KES 3 per page · never expires
- ✓2,000 extraction pages
- ✓Use anytime — no monthly reset
- ✓Pay with Paystack (M-Pesa, card, bank)
Enterprise
Custom volume & negotiated rates
For SACCOs, MFIs, and banks that need large credit pools, invoicing, SLA, and template onboarding. We price to your portfolio — not a monthly subscription.
- ✓Custom credit volume
- ✓Negotiated per-page rates
- ✓SLA & dedicated support
- ✓Multi-branch setup
Add-ons
Unlock underwriting workflows on top of your credits.
Credit review
Repayment capacity, M-PESA + bank combine, underwriting signals, and PDF reports.
Personal analysis
Spending by category, wellbeing score, and actionable tips toward financial freedom.
APIs for developers
Integrate once. Extract everywhere.
REST API for extraction, webhooks when runs complete, and structured JSON for every document type. Wire Textrion into your LOS without rebuilding parsers.
{
"event": "extraction.completed",
"run_id": "a1b2c3d4",
"document_type": "mpesa_statement",
"integrity_grade": "good",
"borrower_case_id": "case_8842",
"credit_review": {
"avg_monthly_in": 105000,
"comfortable_installment": [18000, 22000]
}
}Get started
Pilot with your next ten applications
Register free, upload a statement, and walk through integrity checks, credit review, and committee memo on a real file. Scale to cases, WhatsApp intake, and webhooks when your team is ready.
- ✓ Borrower cases with M-PESA + bank coverage
- ✓ Statement integrity grading on every extraction
- ✓ Auto committee memo + credit review PDF
- ✓ CSV, XLSX, JSON, and webhook delivery
Categorized spend, obligations, and Fuliza patterns.
Balances and debits/credits for salaried borrowers.
One repayment picture from mobile + bank runs.