Inspiration Navigating university admissions often forces students to choose between overly rigid, black-box chance calculators and scattered, overwhelming research. Students struggle to understand where they truly stand, how their academic profile compares to target programs globally, and how to keep sensitive high-school records organized securely.

We created Vantage to build a research-oriented university selection workspace centered around honest context rather than false precision. We wanted a tool that combines live web evidence, structured profile analysis, local privacy, and global institutional exploration into a single, cohesive browser dashboard.

What it does Vantage is a private, browser-based workspace designed to help students evaluate target programs, benchmark their progress, and organize high-school records:

Fit Report: Analyzes a student’s academic profile (GPA, test scores, target program, personal context) against target institutions, producing structured summaries, gap analyses, and actionable next steps. Allows side-by-side comparison of up to 3 programs, text export, PDF printing, and local persistence.

Reality Check: Evaluates student standing across grade levels, courses, activities, and notes, providing honest planning reviews alongside historical benchmarks.

Around the World Directory: Offers a multi-tier world directory (continents, countries, provinces/states, institutions) paired with an interactive atlas map and filtering by institution type or city.

High-School Report Vault: A client-side repository allowing students to drag-and-drop transcripts, recommendations, and portfolios. Files are stored entirely inside browser IndexedDB and never uploaded to application servers.

Deterministic Research Simulation: When live research credentials or services are unavailable, Vantage seamlessly falls back to a deterministic, named 24-university global cohort clearly flagged with SIMULATED GLOBAL RESEARCH indicators to guarantee complete offline and demo availability.

How we built it Vantage is built with a modern Next.js App Router stack designed for speed, privacy, and reliable fallback handling:

Frontend & UI: Next.js 13.5 (App Router), React 18, TypeScript (strict mode), Tailwind CSS, Radix UI primitives, Lucide icons, and Recharts for data visual overlays.

Live Research & AI Pipeline: Next.js Server API routes (/api/fit, /api/reality-check) integrate SerpApi for live web scraping of admissions requirements and Groq for rapid structured report generation from the gathered evidence.

Offline Fallback Architecture: Custom simulation layer (lib/simulations.ts) providing deterministic research responses across 24 global institutions when API keys are absent or rate-limited.

Privacy-First Local Storage: Browser localStorage for report persistence and IndexedDB (lib/report-store.ts) for zero-server handling of heavy student record uploads.

Deployment: Hosted on Netlify using @netlify/plugin-nextjs for serverless API routing and static asset delivery.

Challenges we ran into Graceful Degradation for Live Research: Third-party web search and LLM APIs can suffer rate limits or downtime. Engineering a transparent fallback mechanism (simulations.ts) that preserves UX flow while explicitly labeling simulated data (simulated: true) required clear state isolation between live and fallback modes.

Client-Side Heavy Storage: Managing file uploads (transcripts, letters, portfolios) strictly inside browser IndexedDB without sending any payload to an application server required building custom browser storage adapters and memory-efficient file retrieval workflows.

Hierarchical Global Mapping: Designing an intuitive breadcrumb and interactive atlas map navigation system that seamlessly translates nested continent, country, state, and institution hierarchy data (data/world.json) into smooth UI transitions.

Accomplishments that we're proud of Complete Privacy Guarantees: Successfully keeping sensitive student records, transcripts, and saved evaluation reports 100% inside the user's browser local storage and IndexedDB.

Uninterrupted Product Demos: Building a robust hybrid research engine that effortlessly bridges live SerpApi + Groq queries with a rich 24-university fallback dataset without crashing or blocking user workflows.

Clean & Accessible Interface: Leveraging Radix UI and Tailwind CSS to craft a clean dashboard layout, complete with multi-program comparison views, interactive map views, and print-ready PDF reports.

What we learned Designing for "Honest Context": Admissions guidance is fundamentally advisory. We learned the importance of framing AI-driven outputs as research assistance rather than definitive admission predictions.

Next.js App Router API Patterns: Fine-tuning serverless route handlers in Next.js 13+ to handle external API orchestration alongside deterministic fallbacks efficiently.

Leveraging Web Storage Standards: Combining localStorage for lightweight state with IndexedDB for raw document storage provides a desktop-class offline experience inside a web browser.

What's next for Vantage Expanded Institutional Coverage: Growing the bundled directory (data/world.json) from a curated set to a broader global database of accredited colleges and universities.

Enhanced Document Parsing: Adding client-side OCR parsing to automatically pull GPAs and course lists directly from uploaded transcripts in the High-School Report Vault.

Multi-Profile Tracking: Allowing counselor or mentor views where students can export encrypted workspace state files to share progress securely.

Expanded Live Research Providers: Adding backup search and model providers to further broaden live query capabilities for niche global programs.

Built With

Share this project:

Updates

Submission history