Natus Vincere (Na’Vi)
The Users Employers / Immediate Hirers: Small business owners (café, restaurant, warehouse, retail managers), event organizers, and individuals in urgent need of ad-hoc staff (e.g., emergency dishwashers, movers, delivery couriers, or temporary event helpers). Job Seekers / Gig Workers: University students, part-time workers, and freelancers who have 3–5 spare hours during the day and want to turn that downtime into immediate, same-day earnings. The Problem Slow, Bureaucratic Hiring Processes: Traditional recruitment boards require full CVs, cover letters, and multi-day approval cycles. This makes them entirely impractical for urgent, short-duration tasks (gigs lasting 2–6 hours). High Friction in Job Posting: Busy employers don't have the time to fill out rigid, multi-step job submission forms when an employee cancels last-minute. They need to post a single sentence and find someone right away. Lack of Real-Time, Location-Aware Alerts: Micro-jobs are time-sensitive. Standard job portals fail to notify local workers instantly, resulting in missed opportunities and empty shifts. Unmonetized Spare Time: Everyday individuals with flexible schedules lack a frictionless, trustworthy platform to discover nearby, immediate micro-tasks and get paid on the spot.
Preliminary score
- Value for the user16.5 / 25
A specific problem and a clear outcome from solving it.
- Prototype and use of AI13.5 / 30
A working core scenario, and what AI actually contributes to the result.
- Quality testing4.5 / 20
Tests, examples of failures, and comparison with the current approach.
- Feasibility4.0 / 15
Data requirements, running costs and a clear next step.
- Originality4.0 / 10
A meaningful difference from common solutions.
What the judges wrote
- Claude Opus 5.5Preliminary39/100
- Strengths
- Your problem framing is concrete: a café manager whose dishwasher cancels last-minute needs help today, not after a multi-day hiring cycle. Your Gemini-based extraction, which turns one Azerbaijani sentence into structured fields (title, pay, time window, phone, category), is a sensible use of AI that cuts posting friction. Your setup walkthrough is detailed and localised, and the Telegram channel broadcast is a practical distribution choice for Baku. You were also honest about the main defect instead of claiming everything works.
- To improve
- Your core marketplace loop is broken: other users cannot see a posted job on the board, only through Telegram. Fixing shared storage, for example with a hosted database, should be the first priority, because the scenario does not work end to end without it. Test the AI extraction systematically with 20–30 real-style messages (slang, missing pay, mixed Azerbaijani/Russian), report field-level accuracy and failures, and compare time-to-post against a normal job board form. Add a feasibility section covering Gemini cost per post, how you will handle worker trust and verification and fake listings, and a concrete pilot such as 10 cafés in one district with a target metric like time-to-fill.
- GPT-6 AstraPreliminary46/100
- Strengths
- Your focus on urgent, short shifts identifies specific users and a concrete hiring problem. Your natural-language extraction, preview step, and Telegram distribution offer a sensible way to reduce posting friction. Your disclosure that other users cannot see published listings is an honest, specific failure finding.
- To improve
- Your first priority should be fixing cross-user listing visibility and verifying the employer-to-worker journey across two independent sessions. Your testing should cover extraction accuracy, missing details, ambiguous dates, and Telegram delivery, while comparing posting time and time-to-first-response with existing hiring channels. Your feasibility plan should estimate model and hosting costs, explain how public contact details and spam are handled, and define a small employer pilot.
The team
Cavidan Mikayılzadə
product
Faxrəddin MəmmədovLead
developer
Uğur Namazov
product