Norina is a desktop app that turns logical sentences — written like in a Word document or composed with drag-and-drop pills — into deterministic automations. No statistical models at runtime: just pure Python, executed on your PC, verifiable line by line.
In real companies, systems don't talk to each other. Today, the gap is bridged by people: they open a PDF, check an ERP, copy a code into a spreadsheet, forward an email. Hours, every day, multiplied across every department.
Norina replaces those hours with a visual flow a non-programmer builds in minutes. The person who knows the rule writes the rule — once.
Pills are words. The learning curve is English, not a new syntax. Execution is pure Python — deterministic and auditable, fit for regulated industries.
The March 2026 UIBM filing covers the architecture that makes the linguistic metaphor viable at production scale.
Norina is a binary you install on the operator's PC. No provisioning, no build step, no cloud keys to manage. Your first flow runs the same afternoon.
# Compiled flow — pure Python, readable by IT from norina import sheet, mail, erp wb = sheet.open("Suppliers_April.xlsx") for row in wb.sheet("april").rows(): if row.amount > 1000: pdf = row.attachment.read_pdf() order = erp.sap.match(row.code) erp.teamsystem.write(order.id, result=pdf.result) mail.send( to=row.owner, subject=f"Result {row.code}", body=pdf.summary(), ) # Run: 0 external calls · 142 rows in 3.4s # Log: 142 operations · 0 errors · audit.pdf generated
The generated Python is readable and signable: IT can audit it, version it, reject it. No opaque execution.