SaaS · Full-stack

Developer Tools SaaS

A productivity platform that helps developers manage workflows, automate repetitive tasks, and keep project context in one place — built to reduce context-switching and ship faster.

Role

Lead Engineer

Type

SaaS Product

Duration

8 months

Status

Live

Developers spend a surprising amount of time not writing code — switching between tools, hunting for context, re-explaining tickets, and manually tracking what changed. This platform was built to solve that.

I led the full-stack development end to end: from architecture decisions and data modelling, through to the React frontend and a GraphQL API layer serving real-time updates. The product was designed for small-to-medium engineering teams working asynchronously.

The Challenge

Existing tools were either too heavy (Jira, Notion) or too simple (plain to-do lists). Teams needed something that connected code context — commits, PRs, branches — with task management without a migration nightmare.

The Solution

Built a lightweight layer that hooks into GitHub via webhooks, infers task relationships from commit messages, and presents developers with an auto-updated board — requiring zero manual updates from the team.

~40%

Reduction in status update meetings reported by beta teams

Faster onboarding for new team members joining mid-sprint

500+

Tasks tracked across pilot teams in the first month

React + TypeScriptFrontend UI & state management
Node.js + ExpressAPI server & webhook handling
GraphQL (Apollo)Flexible data querying & subscriptions
PostgreSQLPrimary relational data store
RedisSession caching & real-time pub/sub
Docker + AWS ECSContainerised deployment
01

GraphQL subscriptions are powerful but need careful rate-limiting at scale — learned this the hard way when a spike in webhook events caused client storms.

02

Designing a good data model for "flexible task relationships" upfront saved months of painful migrations later. Schema design deserves more time than most projects give it.

03

The best developer tools are invisible. The more we reduced required clicks per action, the higher the daily active usage climbed.