🎓 Education Guardian Agent Inspiration Applying to U.S. colleges as a student from Nigeria was one of the hardest challenges I’ve ever faced. With limited access to reliable information, confusing requirements like the CSS Profile, and slow support from overstretched advising centers, I had to piece things together alone. I was lucky—eventually, I found EducationUSA Abuja and got into Wesleyan University. But many students aren’t as fortunate. A fellow Wesleyan African Scholar from Sudan shared how he completed his application from a refugee camp, hitchhiking to another country to sit for the SAT—all using just a mobile phone. That’s what inspired Education Guardian Agent—a personalized, AI-powered college counseling assistant that gives every international student access to guidance, no matter their background or location.
What It Does EducationGuardianAgent is a multi-agent system that acts as an intelligent college admissions coach. It collects the user's academic history, goals, preferences, and challenges, and routes them to a network of specialized agents to provide tailored, real-time guidance.
💡 Core Sub-Agents:
Goal Setting Agent – Understands student background and helps define application goals University Matching Agent – Suggests reach, match, and safety schools based on GPA, region, financial constraints, and interests Test Prep Agent – Builds a personalized SAT/ACT study plan with resource suggestions Essay Mentor Agent – Analyzes the user’s story and generates personalized essay topic suggestions, outlines, and feedback Extracurricular Agent – Recommends activities that align with personal interests and target majors Recommendation Agent – Offers advice on choosing and approaching ideal recommenders
Each agent contributes insights that are aggregated into a clear, actionable roadmap for the user’s college application journey.
How We Built It Framework: Built using Google’s Agent Development Kit (ADK) Architecture: Root agent handles intake, context management, and orchestrates interactions across all sub-agents Session Management: Uses InMemorySessionService to maintain persistent user state and interaction history Contextual Reasoning: Pulls from user academic, extracurricular, financial, and personal background using utility tools Essay Agent: Designed a full pipeline for topic generation, profile-based analysis, and prompt-specific drafting support
Frontend: Integrated the agent system with a web UI for an interactive and user-friendly chat experience
Challenges We Ran Into Designing smooth agent handoffs that preserved user context across multiple specialized flows Structuring complex nested user profiles and ensuring every sub-agent could access only the data they needed Handling edge cases like low test scores or missing data, and still offering meaningful, compassionate guidance
Accomplishments We’re Proud Of Simulated a complete college advising experience using conversational AI—entirely automated and highly personalized Built a working end-to-end system capable of interacting with real students, identifying blockers, and offering solutions Successfully replicated the guidance of real counselors across essays, test prep, and school selection
What We Learned AI can’t replace empathy, but it can scale access to critical support where human resources are limited Modular design through Google ADK allows for powerful orchestration between independently intelligent agents
What’s Next Add voice and mobile accessibility for students in low-bandwidth regions Translate into French, Arabic, and other common languages for African and Middle Eastern students Integrate AI feedback loops to improve recommendations based on success outcomes Partner with real organizations (like EducationUSA) to deploy in underserved regions
Log in or sign up for Devpost to join the conversation.