Klyro
The User Entry-level tech professionals (students, recent graduates, or junior developers) aiming to land their first internship or junior role in tech (Frontend, Backend, Data, etc.) The Problem Skill Gap Blindness: They apply to jobs and get rejected without feedback. They don't know exactly which skills they are missing or failing to showcase in their CV for their target role. Overwhelming Roadmaps: Existing online roadmaps are too massive and generic. Users don't know what to focus on right now based on their current level. Lack of Interview Practice: They struggle to articulate their technical knowledge and lack a safe environment to practice and receive objective feedback before real interviews.
Preliminary score
- Value for the user14.0 / 25
A specific problem and a clear outcome from solving it.
- Prototype and use of AI7.5 / 30
A working core scenario, and what AI actually contributes to the result.
- Quality testing3.5 / 20
Tests, examples of failures, and comparison with the current approach.
- Feasibility4.5 / 15
Data requirements, running costs and a clear next step.
- Originality3.5 / 10
A meaningful difference from common solutions.
What the judges wrote
- GPT-6 AstraPreliminary36/100
- Strengths
- Your focus on first-time tech applicants connects specific problems—unclear skill gaps, overwhelming roadmaps, and limited interview practice—into one coherent journey. Your explicit 70/20/10 matching formula makes the proposed recommendations explainable. You clearly distinguish the demo-data frontend and rule-based matching from the planned LLM capabilities.
- To improve
- Your current mock-data flow does not demonstrate GPT-4o contributing to results; prioritize one integrated path from real student input to useful, AI-generated feedback. Your testing report should include reproducible formatting and server-crash cases, fallback behavior, and comparison with manual career preparation. Your feasibility plan needs named internship-data sources, CV privacy handling, and estimated model and hosting costs per session. For your enterprise track, identify an organizational buyer and workflow, then validate an adoption metric such as adviser time saved per student.
- Claude Opus 5.5Preliminary30/100
- Strengths
- Your user is real and well framed: entry-level tech candidates who get rejected without feedback and drown in generic roadmaps. Your matching logic is transparent and explainable (70% required skills, 20% preferred, 10% role alignment), and the deck is unusually honest that internship matching is rule-based, not AI. The five-step flow from role selection to mock interview is coherent, and the Azerbaijani-language interface is a thoughtful local touch.
- To improve
- By your own account, the AI does not yet run: the demo uses mock data and rule-based logic, and LLM analysis, CV parsing and interview evaluation are planned. Getting even one real GPT-4o step working end to end, such as gap analysis from a pasted CV or scoring one interview answer, would change how this is judged. Your testing notes describe formatting issues and server crashes but give no concrete cases, so log specific inputs, outputs and failures, and compare your gap analysis with what a career advisor or a standard roadmap would recommend. You also submitted to the Enterprise track without naming a business buyer, adoption metric, data source or running cost; framing this for universities or HR teams with a measurable outcome would strengthen it.
The team
Şirağa Murtuzayev
data
Ramin SüleymanovLead
developer