Personalized Career Portal
A private career-search workspace that turns saved opportunities and candidate-approved evidence into organized, editable application workflows.
Job searches become fragmented quickly: promising roles live in browser tabs, resume variants lose their context, application answers are rewritten from scratch, and research or networking notes sit elsewhere. I conceived the Personalized Career Portal as a private operating system for a focused search built around a small set of high-fit opportunities rather than hundreds of weak matches.
The application brings opportunity tracking and application writing into one workspace. A candidate can import supported public job postings or enter a role manually, review the extracted details, and move it through a pipeline. The command center summarizes those persisted opportunities. Each application can hold multiple hiring-manager questions and a cover letter, with editable drafts, version history, word limits, and reference documents. AI assistance is optional and grounded in the selected job, approved context, and uploaded PDF, DOCX, text, or Markdown material. Workspace owners can configure encrypted Anthropic or OpenAI credentials and switch providers without exposing keys to other members.
I shaped the product scope, workflows, safety boundaries, and system architecture, then used agent-assisted development to implement and test the full-stack pilot. It is built with Next.js, Supabase Auth/Postgres/Storage, row-level security, and Vercel. Workspaces use owner, editor, and viewer roles so trusted family members can collaborate with different permissions.
The project is live as an internal pilot. Authentication, opportunity management, document-backed application drafting, credential handling, and privacy controls are implemented. The interface also names future or data-dependent areas—such as job discovery, people research, outreach, resume variants, deep research, and story preparation—but these show honest empty states in connected mode until real records or provider integrations exist. It does not scrape LinkedIn or send outreach automatically.
Gallery
Command center populated with the application's synthetic demonstration opportunities.
Full resolution : Career portal command center showing opportunity, application, networking, and fit-scoring summaries alongside a prioritized pipeline.
Opportunity tracker in local demo mode, with synthetic roles and lifecycle controls.
Full resolution : Opportunity tracker listing synthetic job opportunities with stages, fit scores, next steps, due dates, and edit, archive, and delete controls.
Application-writing workspace using illustrative questions and no personal documents.
Full resolution : Application workspace with question selection, an editable answer area, model selection, reference upload, and iterative AI steering controls.Deep dives
No published articles yet.