LOSANI HOMES • ENTERPRISE B2B • 2025
Community microsites without engineering bottlenecks
- Role
- Product Designer
- Timeline
- 2024 – 2025
- Team
- Losani + engineering partners
- Skills
- Enterprise product systems, Cross-functional leadership with PM and Eng, Systems thinking and design systems
Live prototype
Problem
Losani needed community-specific web experiences that could launch at different lifecycle stages — future land, coming soon, available now — while marketing teams updated copy and pricing without waiting on engineering for every change.
The product surface spans buyers, sales, and admin workflows. Complexity comes from multi-role needs, not from visual decoration.
Process
I designed BuiltOps as a modular platform: Next.js 16, Tailwind CSS 4, Payload CMS, TypeScript — SQLite locally, Postgres in production.
Templates encode stage logic in code (safety rails) while editorial content stays flexible in the CMS. Empty sections hide automatically so half-finished pages never ship broken.
Deliverables include community pages, model pages, a version archive, preview-version widget catalog, and lot editor tooling — coordinated with the Losani Connected Site integration.
Outcome
Working platform integrated into the Losani connected-site repo with PR #7 — preview deploy live, production alias pending admin handoff.
Marketing can iterate copy and media in Payload while design maintains layout integrity and stage rules in code.
What I'd do differently
Lock CMS field schemas with design earlier — a few retrofits were needed when marketing wanted section variants we hadn't modeled.
My role
Owned
- Platform UX for community and model pages
- Editorial vs. engineering boundary (Payload CMS vs. layout code)
- Stage-aware page templates (Future / Coming Soon / Available Now)
- Preview, version archive, and draft workflows
Did not own
- Losani corporate brand governance
- Production infrastructure and Vercel team admin
- Sales CRM integrations
Collaborators: Elevow and Losani Homes teams
Key decisions
Decision 1
Hardcoded pages vs. CMS-owned content
Chose: Payload owns copy, images, prices, links, section order, users, and version history. Code owns layout, design, hide-if-empty rules, and stage safety rails.
Rejected: Fully hardcoded marketing pages — fast initially, but every copy change became an engineering ticket and blocked the marketing team.
Decision 2
Single site vs. modular subpages
Chose: BuiltOps as a vendored module inside the connected-site repo — community subpages proxied through the portal origin so buyers stay in one experience.
Rejected: Standalone microsites per community with no shared system. Unmaintainable at scale.
Decision 3
Preview fidelity
Chose: Draft preview + `/old` version archive with frozen playgrounds so stakeholders can compare iterations without risking production.
Rejected: Screenshot-based review only. Stakeholders couldn't validate responsive behavior or CMS-driven edge cases.