Inspiration

The idea for OkEsha came from a personal journey of realizing that education often pushes people to study for marks, degrees, and salaries rather than for knowledge. I saw students graduating without clarity, professionals feeling stuck, and founders lacking the skills to execute their ideas. Traditional career guidance is fragmented, outdated, and inaccessible. I wanted to build something that gives people the clarity I never had: a system that can analyze their goals, understand their strengths, and guide them through skills, projects, and real-world execution. That became the inspiration for OkEsha: an AI agent that empowers knowledge, not just outcomes.

What it does

OkEsha is an AI-powered career and skill advisor that acts as a personal mentor. It helps users understand what to learn, which path to take, and how to build real-world projects that develop job-ready skills.

Key capabilities:

Generates personalized career pathways based on user background

Recommends skills, learning resources, and relevant tools

Creates project ideas and converts them into structured project plans

Suggests project-building tools, tech stacks, and workflows

Acts as an AI mentor to answer questions, guide decisions, and track progress

In a single click, users move from confusion to a clear, actionable roadmap.

How we built it

OkEsha combines multiple layers of intelligence and engineering:

AI Layer: OpenAI models for reasoning, skill mapping, project generation, and multi-step agentic actions

Backend: Supabase for authentication, user profiles, project data, and knowledge mappings

Frontend: Built with Kiro for rapid iteration, clean UI, and modular components

Integrations: Automated workflows and structured prompts for generating paths, projects, tools, and resources

Architecture: A lightweight agentic system that orchestrates user input, data retrieval, and AI reasoning to deliver real-time personalized guidance

The system is designed to be fast, scalable, and extensible for future automation.

Challenges we ran into

Ensuring AI outputs remain structured, stable, and context-aware

Mapping user data to dynamic AI pathways without inconsistency

Handling multi-step reasoning without losing accuracy

Creating a seamless user experience within a short hackathon timeline

Balancing complexity of recommendations with simplicity of UI

Integrating backend tables for projects, planning data, and tool suggestions

Designing prompts that reliably generate actionable project plans

Each challenge forced iterative testing, refinement, and rethinking of architecture.

Accomplishments that we're proud of

Built a fully functional AI agentic career advisor within the hackathon window

Designed a complete workflow from user onboarding to skill recommendations and project planning

Achieved stable AI generation for project ideas, career paths, and tool stacks

Integrated backend storage that maps all user choices to personalized outputs

Created a simple, intuitive interface that makes complex AI guidance feel effortless

Crafted a vision that can truly scale to help millions

What we learned

How to design and refine multi-step AI reasoning flows

How to structure prompts to achieve consistent, production-grade outputs

How to architect a real-time agent system with minimal latency

How to think in terms of user clarity, not just technical features

How to convert vague goals into quantifiable learning paths and project plans

The importance of blending UX, AI, and backend design from day one

What's next for OkEsha

The road ahead is ambitious and clear:

Multi-agent system for deeper career and skill analysis

Voice-based AI guidance for hands-free mentoring

Integration with job platforms for real-time skill-to-job mapping

A full project execution suite for task management, resources, and budgeting

Personalized dashboards with daily skill targets and progress scores

Community-driven learning groups and team-based projects

College and enterprise onboarding for large-scale impact

OkEsha’s long-term vision is to reshape how people learn, build, and grow. Our mission is simple: empower people's knowledge and create pathways that unlock real potential for millions.

Built With

Share this project:

Updates