billing and proposals at nagaris
Nagaris. Sole frontend engineer, November 2025 to March 2026.
Sole frontend engineer on a practice management and billing platform for accounting firms. Built the invoicing, payments and proposal surfaces from nothing.

the problem
When I joined, the product had no billing UI at all. Accounting firms needed to invoice, track payments, schedule them, and price their services through proposals with tiered payment options, and none of that existed on the frontend. With no other frontend engineers on the team, I had full ownership over architecture, UI patterns and how the product felt to use day to day.
invoices and payments
My biggest contribution was the Invoices and Payments experience, built from scratch: the full invoicing surface, payment tracking, payment schedules and the Xero integration, plus the navigation around it. That meant the payment infrastructure on the frontend side too, with card and bank account flows, Direct Debit Service Agreement compliance, and a client onboarding flow with validation checks before payment setup.
People don’t tolerate bugs when money is involved, so I spent more time on edge cases and error states here than on entire features elsewhere.
engagements and proposals
The Engagement and Proposal system is how firms price their services, build proposals with tiered payment options and send them to clients for acceptance. The data model is deeply nested, engagements have phases, phases have deliverables, deliverables have line items, so most of the effort went into making that not feel overwhelming in the UI. The components were built to be reused across the client-facing and accountant-facing sides, so the backend team could drop them into their own views without duplicating work.
communications and consistency
Later came a unified communications inbox for email and SMS, with compose, threading and per-contact actions, rolled out behind feature flags in PostHog. Beyond individual features, I owned frontend consistency across the product: visual standards, Australian date localisation, interaction patterns, and integration quality across Xero, XPM, Nylas, Monoova, PostHog and SendGrid.
what it proved
A smaller company meant speed: less process, less overhead, just shipping quality code. AI tools were even more valuable as a solo frontend engineer, keeping the pace up and leaving more energy for design decisions and less for repetitive implementation. Nagaris proved that real impact is possible at a large company or on a small product team, and gave me confidence in adapting, taking ownership and delivering across very different environments.