MARKRAHIMI.
All projects

2023 — 2025

Gozaresha

An internal reporting tool that grew into a SaaS product with paying customers

Gozaresha

Overview

Gozaresha started as an internal reporting tool. Under my technical leadership, it grew into a multi-tenant SaaS product within months of its relaunch. It handles dynamic form creation and permission-scoped reporting at a scale where correctness actually matters to paying customers.

I designed the data layer on MongoDB, built the auth and permissions system, and led a team of five developers through the transition. We ran Scrum with proper tooling and regular ceremonies, and shipped through Git with Docker for consistent, repeatable deploys.

Turning an internal tool into a product forced a different engineering posture. Every schema change needed a migration path for live tenants. Every deploy needed to happen without downtime for customers who already relied on the system.

The problem

An internal reporting tool, built to replace spreadsheets, had outgrown its original scope. It had none of the multi-tenancy, permissions, or reliability guarantees a paying customer expects from a real product.

The approach

A rebuilt auth and permissions model on MongoDB, a Next.js frontend for the reporting and form flows, and a Docker-based deploy pipeline through Git that let the team ship to multiple environments automatically. Deadline accuracy rose to about 88% under a proper Scrum cadence.

Key features

  • A dynamic, permission-scoped form and report builder tied into existing accounting and project-management systems
  • Auth and role-based permissions built for paying customers, not just internal staff
  • MongoDB as the data layer behind reporting, forms, and notifications
  • Docker and Git-based deploys for repeatable releases across environments
  • A Scrum delivery process with project-management tooling that raised the team's deadline accuracy to roughly 88%

Impact

20+

Companies using the product

5

Developers led through the SaaS transition

88%

Deadline accuracy under the new process

0

Downtime deploys after the Docker rollout

Have something similar in mind?

I'm open to software, full-stack, and AI engineering roles and projects across France and Europe.

Get in touch