Navigos Talent One is the employer recruitment platform that Navigos Group's clients use to post jobs, screen candidates with AI scoring, build hiring campaign pages, and track applications from search to onboarding. Four engineering squads ship features on this codebase in parallel.
203K
Lines of code
15
Packages
4
Squads
2800+
Components
Four squads work in a 203K-line TypeScript monorepo with 15 packages. Changes to shared state could break another squad's feature, and shared files caused frequent merge conflicts. Translation keys were unchecked strings, so a typo could leave a missing label in production. We needed module-level state alongside shared permissions, authentication, and user preferences, plus translation keys that TypeScript could check.
Application tracking with AI match scoring
I built the state system with React's useSyncExternalStore and Immer. Each module has its own scope; app-wide stores hold shared data such as permissions and authentication. Components subscribe to specific values and re-render when those values change, reducing the effect of one module's updates on the rest of the page.
Translations use plain TypeScript objects. Developers get key autocomplete, and the compiler catches misspelled keys without a separate translation build step. Feature modules load on demand, with route prefetching to prepare upcoming pages. Git submodules separate the squads' feature scopes and reduced conflicts over shared files.
Block-based landing page builder
Job management with live metrics
AI-powered candidate recommendations
Candidate AI evaluation profile
The job description generator turns a title and a few bullet points into a structured draft for the employer to review. AI also scores incoming applications across skills, experience, and education, giving recruiters a match percentage alongside each candidate.
Employers can build recruitment pages with a block editor, templates, theme colors, responsive previews, and draft and publish controls. Uploaded images are converted to WebP and resized for different breakpoints to reduce download size. These pages previously needed a developer to build and deploy them.
The ATS combines pipeline stages, AI match scores, bulk actions, and shared reviews. When a recruiter moves a candidate, the interface updates while the server confirms the change. Four squads built parts of this system in parallel, using the shared state architecture to keep module changes separate.
Each squad owns its features and can deploy independently. Login, language, and role changes update the affected components without a full page reload. Route prefetching prepares upcoming pages, modules load on demand, and optimistic updates let the interface respond while a request is in progress. All modules use the same permission checks.
0
Engineering squads
Each ships features without waiting on the others
0K
Lines of TypeScript
Spread across 15 packages in one monorepo
0
Page reloads needed
Login, role, and language changes update in place
0
Shared permission system
Every module uses the same access control logic
