Distribution businesses run on sales updates posted into WhatsApp groups and a reconciliation spreadsheet at the end of the month. Nobody can see what reps are actually selling, which clients have gone quiet, or whether stock runs out before it becomes a problem.
Success looks simple: a manager catches a dormant client or a stock-out before it costs revenue, instead of finding out at month end.
Reps file a daily sales report from the field. Managers approve or flag it, so a problem surfaces the same day. Owners read one of 11 role-specific dashboards, each customisable by drag and drop, backed by materialised views so they stay fast as the data grows.
Around that sit invoicing, inventory and billing, plus background workers that do the chasing nobody remembers to do: overdue invoices, payment reminders, inventory demand prediction, dormant-client detection and anomaly detection.
An AI assistant answers questions about the business’s own data, and which model powers it is configurable per tenant: Claude, GPT, Gemini or a self-hosted model for anyone who will not send data to a third party.
Every tenant-scoped query runs through a single guarded path that injects the organisation, and that isolation is covered by automated tests rather than assumed.
Financial identifiers such as NTN and STRN numbers, bank accounts and IBANs are encrypted with AES-256-GCM transparently at the database layer, so the application reads plaintext while the database only ever stores ciphertext.
Built on Next.js and PostgreSQL, with English shipped and the structure already in place for Urdu, right-to-left included.