-
-
Landing page showcasing platform features, AI capabilities, student dashboard, notifications and personalized actions
-
Profile Creation for AI-Powered Roadmaps
-
Multi-level AI processing creating personalized learning pathways
-
AI Mentor for Daily Learning and Real-time educational support with live citations and resources
-
24/7 personalized career guidance with real-time web intelligence
Inspiration
1. Problem Statement
1.1 The Current Educational Crisis
In an era where Artificial Intelligence is revolutionizing industries and redefining career paths, a vast majority of students — particularly from Tier 3 institutions, rural and semi-urban colleges — are still left behind. The root of the problem isn't their lack of talent or interest, but the lack of access to updated resources, structured mentorship, and personalized career direction.
1.2 Institutional Limitations
Most institutions still follow outdated curriculums, operate on paper-based systems, and lack basic digital infrastructure. A single teacher managing 100+ students cannot realistically understand each student's individual strengths, struggles, or aspirations. As a result, students either receive generic, one-size-fits-all advice or are left unguided, unsure of their path in the competitive job market.
1.3 The Employability Divide
Meanwhile, companies and recruiters often assume that only top-tier college students possess strong skills — not always because it's true, but because those students have access to better support systems, exposure, and grooming. This assumption creates a deep divide in perceived employability, pushing countless capable students into confusion or underconfidence.
1.4 The Information Paradox
While the internet today offers thousands of free courses, open-source projects, and career content, the real challenge is not content availability, but lack of personalized guidance. What works for one student might not work for another. Every student's journey should be mapped based on their unique interests, background, strengths, and evolving industry demands.
1.5 The Mental Health and Hidden Potential Crisis
The Mental Health Impact in This Competitive World:
In today's hyper-competitive environment, students are facing unprecedented levels of stress, anxiety, and depression. The constant pressure to succeed, combined with lack of proper guidance, has created a generation struggling with mental health issues.
- Students develop depression and anxiety from constant comparison with others
- Imposter syndrome makes them believe they're not capable of success
- Fear of judgment prevents them from exploring their true potential
- Many students lose hope and accept that success is only for others
The Hidden Potential Crisis:
Every student has unique strengths and talents, but without proper guidance, these remain undiscovered. Students need supportive environments where they can:
- Discover their natural abilities without judgment
- Explore career paths that align with their interests
- Build confidence through personalized guidance and encouragement
- Learn in a fun and engaging way that reduces stress and anxiety
With proper guidance and support, every student can become successful. The key is making learning fun, happy, and personalized to help students unlock their true potential and build the confidence needed to pursue their dreams.
2. Our Solution
2.1 PBSC-Ignite: The Core Idea
The Problem We Solve: Students aren't lacking information — the internet is flooded with courses, tutorials, and career content. What they're missing is intelligent guidance on what to learn, when to learn it, and how it fits their unique strengths and career goals.
Our Approach: PBSC-Ignite is an AI-powered career navigator that doesn't create new content but makes existing resources work smarter for each student. We analyze their LinkedIn and GitHub profiles to understand their current skills, identify their natural strengths, and map personalized learning paths using the best available resources on the internet.
What Makes Us Different:
- We don't replace teachers — we enhance their capabilities
- We don't compete with colleges — we strengthen their offerings
- We don't create courses — we intelligently curate existing ones
- We don't give generic advice — every recommendation is personalized to individual strengths
2.2 How It Works (Complete Learning Cycle)
- Student connects their LinkedIn and GitHub profiles
- AI analyzes their skills, projects, and career goals
- System creates a personalized roadmap using existing online resources
- Learning modules are divided into manageable weekly and daily tasks
- AI continuously monitors progress and completion of each module
- AI evaluation system assesses understanding and skill development
- Leo AI provides ongoing career guidance with real-time industry insights
- Platform adapts future recommendations based on learning performance
2.3 The Complete Learning Intelligence System
Module-Based Learning:
- Structured Phases: Career roadmaps divided into logical learning phases
- Weekly Modules: Each phase broken into manageable weekly learning blocks
- Daily Tasks: Specific, actionable tasks with clear completion criteria
AI-Powered Progress Monitoring:
- Completion Tracking: System monitors which modules students have finished
- Skill Assessment: AI evaluates understanding through various methods
- Learning Pattern Analysis: Understanding how each student learns best
- Adaptive Recommendations: Future modules adjusted based on performance
Intelligent Evaluation:
- Non-Traditional Assessment: Moving beyond simple tests to practical skill evaluation
- Portfolio Analysis: Reviewing GitHub commits, projects, and LinkedIn updates
- Peer Comparison: Benchmarking against industry standards and successful peers
- Continuous Feedback: Ongoing suggestions for improvement rather than one-time grades
2.4 Current Implementation & Future Vision
What We've Built (Student-Focused Platform):
- Personalized career roadmaps based on profile analysis
- Module-based learning with progress tracking
- AI-powered career coach (Leo) for real-time guidance
- Smart resource discovery and curation
- Learning completion monitoring system
Future Extensions We Can Build:
For Educators:
- Classroom Analytics Dashboard showing student progress across modules
- AI Evaluation Insights helping teachers understand individual student performance
- Industry Alignment Tools keeping faculty updated on relevant skills
- Student Support Alerts identifying students who need additional help
For Institutions:
- Career Services Enhancement Platform with comprehensive student tracking
- Placement Success Analytics correlating learning completion with job outcomes
- Curriculum Optimization based on AI evaluation of student performance
- Alumni Success Monitoring for continuous program improvement
2.5 Supporting UN SDG 4: Quality Education
PBSC-Ignite directly aligns with UN Sustainable Development Goal 4 by ensuring "inclusive and equitable quality education and lifelong learning opportunities for all." Our platform democratizes access to personalized career guidance, making quality educational support available to every student regardless of their institutional background or resources.
The Bottom Line: We're building a complete intelligent learning ecosystem that guides, monitors, evaluates, and continuously improves the learning experience for everyone — ensuring no student is left behind and every individual can reach their full potential.
3. Our MVP Implementation
PBSC-Ignite transforms career guidance through intelligent orchestration of cutting-edge AI technologies. Our platform addresses the fundamental challenge of providing current, accurate career advice in a rapidly evolving job market where traditional LLMs, while powerful, are limited by their training data cutoffs and can produce hallucinations when asked about current industry trends.
At the core of our innovation is the strategic use of Perplexity Sonar's real-time web intelligence capabilities. Unlike traditional LLMs that rely on static training data, Perplexity provides live, cited information directly from current web sources, eliminating hallucinations and ensuring accuracy. We've implemented a 3-Level Enhancement Pipeline where Perplexity serves as our "industry observer" - accessing real-time job market data, current learning methodologies, and the latest educational resources with proper citations and sources.
However, while Perplexity excels at observation and information retrieval, we needed advanced reasoning capabilities to structure this raw intelligence into personalized, actionable guidance. This is where our dual-layer orchestration becomes powerful - Perplexity's factual, cited observations are processed through Groq's high-speed LLaMA reasoning engine, which excels at understanding context, logical structuring, and personalized adaptation. Groq's infrastructure provides the computational speed needed for real-time reasoning while LLaMA's advanced language capabilities ensure the guidance is coherent, personalized, and pedagogically sound.
Our system combines the best of both worlds: Perplexity's real-time, cited accuracy with LLaMA's sophisticated reasoning and personalization capabilities. We've built a comprehensive career navigation ecosystem that analyzes students' LinkedIn and GitHub profiles, uses vector databases for intelligent context retrieval, and employs smart caching strategies for optimal performance. The result is guidance that is both factually current and intelligently personalized.
3.1 Technology Stack
Here are the core technologies powering our AI-driven career guidance platform:
3.1.1 Core AI Layer
- Real-time Intelligence: Perplexity Sonar Pro
- Advanced Reasoning: Groq LLaMA 4 Scout
- Adaptive Learning Intelligence: Amazon Bedrock Claude Sonnet 4
3.1.2 Backend Infrastructure
- Framework: Flask (Python)
- Database: MongoDB
- Authentication: bcrypt
- High-Performance Caching: Redis
- Caching: 15-day MongoDB caching
3.1.3 External Integrations
- AI Services: Perplexity API, Groq Cloud API
- Professional Data: LinkedIn Data API, GitHub REST API
- Learning Resources: YouTube API, Google Scholar API, Google Custom Search API
- Content: Medium API
3.1.4 Frontend Layer
- HTML, CSS, JavaScript, Bootstrap – For a responsive and intuitive user interface.
3.2 Student Home Dashboard & Platform Overview
3.2.1 Interactive Landing Page & Core Features Overview
Once a student creates an account, they are welcomed with an interactive landing page that serves as a dashboard, offering a brief demo of the platform's key features and available resources. Apart from the core AI-powered roadmap and mentorship, the platform also provides several additional functionalities:
Tech News & Articles: Users can stay updated with the latest developments through curated content from Medium API with category-based browsing and pagination support.
Company Timeline: Display of top 5 companies that visited for campus recruitment with visit dates and role offerings.
Navigation Hub: Seamless access to profile management, career roadmaps, Leo AI coaching, AI tutoring, and learning resources.
The dashboard employs responsive design using Bootstrap 5 and JavaScript for dynamic interactions, ensuring optimal user experience across all devices.
3.3 Profile Creation for AI-Powered Roadmaps
3.3.1 Why Profile Creation Matters
Before generating an AI-powered learning roadmap, the system must first understand the user holistically. The profile creation process involves collecting user interests, career goals, and academic preferences, gathering information about dream companies, desired job roles, and industry sectors, and most importantly, linking to social profiles like LinkedIn and GitHub for a personalized AI-driven experience.
The system analyzes career aspirations, preferred work environments, and conducts intelligent assessment of strengths and weaknesses based on actual professional and technical presence rather than just stated preferences.
3.3.2 Data Collection & Processing Pipeline
Once profile data is collected, the system initiates a comprehensive data extraction process. LinkedIn Data API provides professional background including headline, summary, skills, certifications, work experience, and network connections. GitHub REST API analyzes repositories, coding patterns, project complexity, technology preferences, and contribution history.
Smart Data Storage Strategy:
- MongoDB Users Collection: Stores basic profile information, career goals, preferences, and cache timestamps
- MongoDB LinkedIn Collection: Stores extracted professional data with 15-day TTL for cost optimization
- LlamaIndex Document Processing: Transforms raw profile data into structured, intelligent documents
- Qdrant Vector Database: Stores vectorized profile documents for semantic search and enhanced AI context
The system employs intelligent 15-day caching to optimize API usage - checking cache timestamps before making external calls, reducing operational costs by 85% while maintaining data freshness.
3.3.3 Enhanced Context Creation & AI Readiness
The collected data undergoes sophisticated processing using LlamaIndex framework to create rich, contextual documents that capture not just what students have done, but insights about their learning patterns, natural aptitudes, skill gaps, and career potential. These documents are vectorized using HuggingFace embeddings and stored in Qdrant for semantic similarity matching.
Profile Intelligence Ready for AI Processing: The processed profile data creates a comprehensive student context that includes current skill level assessment, career goal alignment, strength identification, and learning preference patterns. This enhanced context becomes the foundation for all AI interactions.
Result: The profile intelligence system now has complete understanding of the student's professional background, technical capabilities, career aspirations, and learning style - creating the perfect input for Perplexity's real-time industry intelligence to generate highly personalized, current, and relevant career guidance.
3.4 AI-Powered Roadmap Generation
3.4.1 How the Roadmap is Built
Utilizing our dual-layer AI orchestration system with Perplexity Sonar Pro and Groq LLaMA 4 Scout, the platform dynamically generates comprehensive learning roadmaps tailored to each user. The system creates a structured journey divided into distinct learning phases, each containing clear timelines and organized content, and delivers detailed weekly learning plans with daily task breakdowns and curated resources.
3.4.2 Detailed Multi-Level Processing Flow
Level 1 - Enhanced Roadmaps Generation: The system sends student profile intelligence to Perplexity Industry Observer for comprehensive career analysis. Perplexity observes current industry trends, analyzes job market demands, and provides intelligent insights about optimal learning paths. This industry intelligence flows to LLaMA Roadmap Reasoner which structures this information into personalized learning phases and stores the initial roadmap in MongoDB.
Level 2 - Detailed Learning Plans Creation: Starting from the generated roadmap phases, the system sends each phase requirements to Perplexity Learning Method Discoverer to find current educational methodologies and effective study approaches. Perplexity returns strategic learning insights and recommended sequences. This methodology data flows to LLaMA Learning Plan Structurer which processes this information with student preferences to generate detailed weekly schedules with daily tasks for each phase.
Level 3 - Personalized Resource Discovery: Starting from the generated weekly learning plans, the system sends each daily learning topic to Perplexity Resource Finder for current educational material discovery. Perplexity identifies the most recent, highly-rated resources with quality assessments. This resource data flows to LLaMA Resource Organizer which personalizes and organizes these materials based on the student's learning context and maps them to specific daily tasks.
3.4.3 Progressive Learning & Milestone System
Daily Task Completion & Adaptive Assessment:
- Each day presents specific learning tasks with auto-generated assessments powered by Claude Sonnet 4
- Students must complete and pass daily tests to unlock the next day's content
- Claude Sonnet 4 continuously adapts difficulty based on student performance and learning patterns
- Redis caching ensures instant progress tracking and seamless user experience
Milestone Achievement & Social Integration:
- Weekly and monthly milestones are automatically tracked based on daily task completion
- Unipile API integration enables automatic LinkedIn milestone sharing once daily tasks are completed
- Students can showcase their learning achievements directly to their professional network
- Progress celebration and motivation through social recognition
3.4.4 Complete Integration & AI Tutor Handoff
After all three levels complete their processing, the comprehensive roadmap data (phases, weekly plans, daily tasks, resources, and adaptive assessments) is integrated and stored in MongoDB. This complete learning intelligence is then prepared for the AI Tutor System, which receives the full context of the student's personalized learning journey for targeted tutoring support.
3.4.5 Results & Systematic AI Orchestration
This multi-level system ensures each AI component specializes in its expertise while building upon previous processing results. The systematic flow from industry observation through learning methodology discovery to resource personalization creates highly effective, current, and personalized learning experiences.
3.5 AI-Powered Learning Tutor
3.5.1 What are AI Tutor Sessions?
The AI-Powered Learning Tutor represents a critical component of the daily learning experience, providing in-depth guidance on specific topics students are studying within their personalized roadmaps. Unlike traditional tutoring systems that rely on static knowledge, our AI Tutor leverages Perplexity Sonar Pro to access real-time, cited information for every topic, ensuring students receive the most current and accurate educational content available.
The AI Tutor creates a specialized learning environment focused exclusively on the current week's learning module, delivering comprehensive conceptual explanations with live sources and citations, and providing multiple learning formats with real-world, up-to-date examples and applications.
3.5.2 Perplexity-Powered Tutoring Architecture
When students access the AI Tutor from their learning plan, the system loads complete module context including learning objectives, key skills being developed, and the student's current progress. The tutoring system is primarily powered by Perplexity Sonar Pro which retrieves live, cited information about the specific learning topic, ensuring all explanations are grounded in current, factual sources rather than potentially outdated training data.
Enhanced Context Integration: The AI Tutor retrieves the student's profile intelligence from Qdrant vector database, understanding their learning style, current skill level, strengths, and preferences. This profile context is combined with the learning module information and sent to Perplexity, enabling highly personalized and contextually relevant tutoring responses.
Fallback System: When Perplexity is unavailable or for supplementary processing, the system employs Groq LLaMA 4 Scout as a fallback to ensure continuous tutoring support, maintaining the same level of personalized guidance.
3.5.3 Real-time Cited Learning Process
The AI Tutor sends the specific learning topic along with student profile context to Perplexity Intelligent Tutor, which retrieves current information about the topic with proper citations and sources. Perplexity provides comprehensive explanations grounded in live web sources, ensuring accuracy and currency of all educational content.
Context-Aware Personalization: For each tutoring interaction, the system:
- Retrieves student profile from Qdrant vector database for personalization context
- Sends learning topic + student context to Perplexity for real-time content discovery
- Receives cited, current explanations tailored to the student's learning level
- Provides sources and references for deeper exploration and verification
Comprehensive Learning Experience: Students receive topic overviews with live citations, detailed explanations with current examples, interactive Q&A with factual, sourced responses, and pathways for deeper learning through provided sources and references.
3.5.4 Dynamic Resource Integration & Results
The tutoring system integrates real-time resource discovery through multiple APIs - YouTube API for current educational videos, Google Scholar API for recent research papers, and Google Search API for up-to-date web resources. All resources are filtered and ranked by relevance to the specific learning module and student's comprehension level.
Adaptive Deep Learning: The AI Tutor enables students to go deeper into topics through cited sources, explore comprehensive coverage of complex concepts, access multiple perspectives on challenging topics, and verify information through provided citations and references.
Current Implementation Results: The Perplexity-powered AI Tutor successfully provides real-time, cited educational content with personalized context awareness. Students report significantly improved trust in the information received, better understanding through current examples and sources, and enhanced learning engagement through the ability to explore and verify information independently.
The AI Tutor creates a dynamic, trustworthy learning environment where every explanation is grounded in current, cited sources while remaining perfectly tailored to individual student needs and learning contexts.
3.6 Career Coach Leo - The Complete AI Hero
3.6.1 What is Career Coach Leo?
Career Coach Leo is PBSC-Ignite's advanced AI career guidance system that serves as students' personal career navigator and strategic advisor. Leo combines real-time web intelligence with deep student understanding to provide personalized, actionable guidance that evolves with current trends and information.
3.6.2 Leo's Intelligence Architecture
Level 1: Student Profile Intelligence Leo retrieves comprehensive student data from Qdrant vector database including career goals, LinkedIn professional data, GitHub projects, learning progress, and personal statements to understand each student's unique trajectory.
Level 2: Real-Time Information Intelligence The system uses Perplexity Sonar Pro to research current information, trends, developments, and insights in real-time across any topic students inquire about, ensuring all responses are based on the most current available data.
Level 3: Personalized Response Creation Leo combines student profiles with live information to create tailored advice, recommendations, and guidance specific to each individual's context and needs.
3.6.3 Complete Student Care System
Comprehensive Monitoring Leo continuously tracks learning progress, skill development, profile changes, and engagement patterns across the entire platform to provide proactive guidance.
Intelligent Notifications
- Learning milestone celebrations and encouragement
- Relevant updates based on student interests and current trends
- Personalized recommendations with live information
- Progress check-ins and motivational support
- Current developments in student's areas of interest
Proactive Intervention Leo identifies struggling students and automatically provides personalized encouragement, action plans, strategy adjustments, and targeted recommendations based on current information.
3.6.4 Leo's Core Capabilities
Real-Time Conversational Intelligence Students can discuss any topic and receive responses grounded in current, live information. Leo fetches the latest data, trends, and insights to provide accurate, up-to-date guidance on whatever students are curious about.
Live Information Research Leo provides current information on any subject by researching real-time data, recent developments, emerging trends, and latest insights with proper citations from Perplexity research.
Personalized Guidance Offers specific recommendations and advice tailored to each student's profile, combining current information with individual context to provide relevant, actionable guidance.
Adaptive Conversations Maintains context across conversations, understanding each student's interests, goals, and previous discussions to provide continuous, meaningful interactions.
3.6.5 24/7 Learning Companion
Complete Learning Ecosystem Support Leo supports students' entire learning journey, providing guidance and information on any topic they're curious about, helping them stay current with developments in their areas of interest.
Emotional Support System Beyond information delivery, Leo celebrates achievements, provides encouragement during challenges, and maintains motivation throughout the learning journey.
Trusted Information Source Creates a reliable environment where students can ask about anything and receive current, cited information combined with personalized guidance that feels both informative and supportive.
3.6.6 Implementation Impact
Students report Leo as their most trusted information companion, providing current insights and personalized guidance on any topic they explore. The combination of real-time information intelligence with personalized understanding helps students stay informed and make confident decisions across all aspects of their learning and development.
Leo ensures students always have access to current, reliable information while receiving guidance tailored to their individual context and needs.
The PBSC-Ignite MVP successfully demonstrates a complete AI-powered career readiness ecosystem that goes beyond traditional learning platforms. Through the integration of Multi-Level Perplexity Enhancement, real-time AI Tutoring, and Career Coach Leo, we have achieved a unified system where students receive personalized roadmaps based on current industry trends, learn with live cited information, and benefit from continuous career guidance.
Perplexity Real-Time Intelligence as System Foundation The breakthrough achievement lies in Perplexity Sonar Pro's real-time web intelligence capability, which fetches live industry data, current educational content, and emerging trends with proper citations. This real-time data fetching plays a crucial role throughout our system - powering dynamic roadmap generation with current market demands, providing AI tutoring with up-to-date educational resources, and enabling Leo to offer career guidance based on live industry conditions rather than outdated static information.
This MVP establishes PBSC-Ignite as a revolutionary platform that combines academic learning with real-world career preparation, delivering measurable results in student engagement, learning outcomes, and career readiness through AI-driven personalization at scale, all anchored by Perplexity's real-time intelligence that keeps our entire ecosystem current and relevant.
4. Conclusion
PBSC-Ignite successfully addresses the critical gap between academic learning and industry requirements by developing a comprehensive AI-powered career readiness platform. Through our MVP development, we have created an integrated ecosystem consisting of personalized roadmap generation, real-time AI tutoring, and continuous career coaching that serves students from profile creation to career preparation.
Key Achievements We developed a multi-level AI processing system that combines student profile intelligence with live industry data to create personalized learning experiences. Our platform integrates LinkedIn professional data, GitHub project portfolios, and comprehensive student profiles stored in vector databases for intelligent matching and personalization. The complete user journey from registration to career readiness is fully operational and tested.
Perplexity Integration Success The integration of Perplexity Sonar Pro as our real-time intelligence foundation has been our breakthrough achievement. This API enables us to fetch current industry trends, live educational content, and up-to-date career information with proper citations, ensuring our platform remains relevant and accurate. Students receive guidance based on current market conditions rather than outdated static information, making PBSC-Ignite significantly more valuable than traditional educational platforms.
Impact and Accessibility Our platform serves UN Sustainable Development Goals by providing quality education and decent work preparation to all communities, including rural and underserved areas. We're building PBSC-Ignite as a powerful tool that anyone can use - every student gets the same quality learning experience no matter where they're from or what their background is.
Future Development Vision Building on our successful Perplexity integration, we plan to expand PBSC-Ignite into a complete educational ecosystem. Future modules will include knowledge graph-based learning structures, comprehensive exam preparation systems, interview coaching modules, and industry-specific certification pathways. Each new component will leverage Perplexity's real-time intelligence to maintain currency and relevance.
PBSC-Ignite demonstrates that intelligent integration of real-time web data with personalized AI systems can create effective, scalable solutions for modern career education challenges while ensuring equitable access to quality learning resources.
5. Live Demo
Experience PBSC-Ignite in action! Watch the complete user journey from profile creation to AI-powered career guidance:
The demo showcases:
- AI-enhanced student dashboard with real-time progress tracking
- Automated profile creation with LinkedIn and GitHub integration
- Multi-level AI roadmap generation powered by Perplexity intelligence
- Real-time AI Tutor providing live citations and educational resources
- Career Coach Leo delivering personalized guidance with industry data
@PBSC-Ignite: Your Personalized Blueprint for Skill and Career by LEO - AI.
Log in or sign up for Devpost to join the conversation.