Inspiration

In today's fast-paced tech world, learners and job seekers often find themselves overwhelmed with scattered resources and unstructured paths. We noticed how difficult it was for individuals—especially students and early professionals—to plan, track, and stay consistent with interview prep or skill-building. Existing platforms were either too generic or lacked true personalization. We wanted to solve this by building a tool that crafts tailored, AI-powered learning journeys that adapt to a person’s goals, background, and available time.

What it does

FocusFlow is an AI-powered personalized learning and interview preparation platform. Users enter their goals (e.g., “Crack Google SDE-1 interview in 30 days”) and optionally upload their resume. FocusFlow then:

  • Generates a day-by-day roadmap based on the goal, timeline, and user's experience.
  • Delivers structured daily study material in rich Markdown—complete with topics, examples, and problem sets.
  • Tracks daily progress and completion status at the topic level.
  • Stores AI chat history to assist with questions on the go.
  • Includes an AI assistant that adapts based on the user’s learning journey and background.

How we built it

We used:

  • Supabase for authentication (Google Login), database, and storage.
  • PostgreSQL as our primary database with a normalized schema to support journeys, topics, content, and progress tracking.
  • Gemini models to generate personalized roadmaps and educational content.
  • A custom AI prompt-engineering layer that considers user goals and resume to generate tailored day-by-day plans.
  • Next.js + Tailwind CSS for the front-end UI.
  • GitHub and Vercel for version control and deployment.
  • Bolt.new for overall project building.

Challenges we ran into

  • Personalization vs Performance: Generating a new roadmap for each unique user prompt while maintaining quick response times was challenging. We needed a hybrid caching mechanism that reused similar roadmaps but adapted them using LLMs for personal nuances.
  • Resume Parsing: Extracting structured data from varied resume formats was tricky and required both rule-based and ML-based approaches.
  • AI Prompt Design: Creating prompts that reliably produced high-quality and varied content without being redundant was an iterative process.
  • Scalability: Ensuring our database schema could support topic-level progress tracking and multi-journey capabilities without overcomplication.

Accomplishments that we're proud of

  • Built a fully functional MVP that delivers personalized, structured learning journeys within seconds.
  • Seamlessly integrated Supabase Auth and DB with role-based access and state persistence.
  • Designed a clean, distraction-free UI focused on content delivery and learning efficiency.
  • Created a flexible AI content generation system that adapts to the user's experience level and study goals.

What we learned

  • The importance of granular progress tracking: Users are more motivated when they can see completion at a micro (topic) level.
  • Effective prompt engineering is critical to getting high-quality AI outputs, especially when personalization is involved.
  • Database-first thinking helped us scale features more systematically as we added chat history, summaries, and more.
  • Leveraging LLMs as planners, teachers, and assistants opens up massive opportunities for personalized education.

What's next for FocusFlow

  • Creator Tools: Allow educators to upload their own roadmaps and content.
  • Community Journeys: Enable users to discover and remix public roadmaps.
  • Progress Analytics: Visual charts to track learning speed and gaps.
  • Offline Mode + Mobile App: Learn and review on the go.
  • Auto-complete Summaries: Let AI generate user notes from content.
  • Company-wise Interview Patterns: Use scraped and verified data to better guide users for company-specific preparation.

FocusFlow aims to become the go-to companion for structured, distraction-free, and deeply personal skill building.

Built With

Share this project:

Updates