Roadmap
What we're planning next.
7 items
Security hardening: rate limiting, headers, password strength
Add rate limiting to all API routes to prevent abuse, set security headers (HSTS, X-Frame-Options, etc.), and increase the minimum password length from 4 to 8 characters to better protect student data.
Teacher-curated opportunity list
Let teachers review, add, and remove opportunities from the curated directory so students see a list vetted by someone they trust — not just AI-researched entries.
Show AI-generated flag on bullets in the builder
Surface a subtle visual indicator on AI-drafted bullets so students always know which parts of their resume were written by AI vs. by them. The flag appears in the builder only — not on the final PDF or printout.
Test and validate all LLM prompts
Systematic testing of all 6 AI prompts — interview questions, bullet generation, bullet refinement, resume parsing, and opportunity search (2 stages). Validating output schemas, edge cases, tone for high schoolers, and hallucination risk.
Enable resume import from existing text
Paste an existing resume and have it parsed into structured sections automatically. Currently disabled while we validate that PII handling in the parser meets our privacy standards.
Final resume review: LLM accuracy check
After all sections are complete, the AI reads the full resume and asks 3–5 targeted questions to catch exaggerations or inaccuracies — like a counselor doing a final read-through before you submit.
Curated opportunity directory
A searchable, filterable list of real volunteer positions, part-time jobs, and programs for Seattle-area teens. The current list was researched with AI (ChatGPT Deep Research). The AI search stays as an optional supplement for personalized results.