Kramo
AI Product
A career decision platform that starts from a resume, a profile paste or a plain description and returns a personalised career map with concrete next actions.
The idea
Kramo — “We Map Your Future” — is a full-stack career decision platform. A user begins with a resume, a pasted profile or a natural-language description of themselves, and receives a personalised dashboard of career directions, each one openable into a deep-dive page with specific next actions.
Most resume tools stop at scoring a document: ATS checks, keyword grades, formatting fixes. That answers the wrong question. The resume is only a starting point — the useful output is direction. Kramo was built so the resume is an input to a decision, not the product itself.
What I built
- Next.js App Router application in TypeScript with a Prisma data layer on Supabase Postgres.
- Resume ingestion and parsing for PDF and DOCX, plus profile-paste and free-text self-description inputs.
- A generated career dashboard of clickable modules, each expanding into its own deep-dive recommendation page.
- A persistent career mentor that reads the latest saved profile and report so conversations keep context.
- Email/password authentication with HTTP-only session cookies, bcrypt hashing and JWT sessions via jose.
- Server-only Gemini integration behind an environment key — no model calls from the browser.
- Account surface: profile, saved reports, settings, logout and account deletion.
- A native Expo / React Native Android client that uses the same backend — native auth, secure session storage and a document picker for resume upload, not a WebView wrapper.
How it works
Dark emerald glassmorphism over frosted sage panels with mint-cyan glow. The flow is deliberately linear: input, then map, then deep dives, then mentor. Loading and error states are explicit at every step so a slow model call never looks like a broken screen.
Visual showcase
