ExoForce
Bakı şəhərində hər gün səhər (08:00–10:00) və axşam (18:00–20:00) pik saatlarında işə və ya universitetə gedən, xüsusilə Nizami–28 May xəttindəki sıxlıqlardan əziyyət çəkən tələbələr, işləyən gənclər və ictimai nəqliyyat istifadəçiləri. Mövcud Problem: Standart xəritə tətbiqləri real vaxtda stansiyaların içindəki basabası, magistral yollardakı probkaları və alternativ xüsusi zolaqlı ekspres avtobusların anlıq vəziyyətini göstərmir. Nəticədə sərnişinlər məlumatsız şəkildə stansiyalarda və tıxaclarda saatlarla vaxt itirirlər. MetroFlow.ai bu problemi həll edərək onlara AI vasitəsilə ən sürətli alternativ yol xəritəsini təqdim edir, canlı xəbərçi modulu ilə isə istifadəçiləri məlumat paylaşmağa həvəsləndirərək EcoPoints və real qəhvə mükafatları (Gloria Jeans) qazandırır.
Preliminary score
- Value for the user15.0 / 25
A specific problem and a clear outcome from solving it.
- Prototype and use of AI10.0 / 30
A working core scenario, and what AI actually contributes to the result.
- Quality testing6.5 / 20
Tests, examples of failures, and comparison with the current approach.
- Feasibility4.0 / 15
Data requirements, running costs and a clear next step.
- Originality5.0 / 10
A meaningful difference from common solutions.
What the judges wrote
- GPT-6 AstraPreliminary44/100
- Strengths
- Your focus on peak-hour commuters along the Nizami–28 May corridor identifies a specific audience and concrete problem. Your reported station-name errors and response delays led to a practical dropdown redesign. Combining congestion reports with EcoPoints offers a plausible incentive for users to contribute local information.
- To improve
- Your materials should demonstrate one complete journey from a timestamped congestion report to an updated route recommendation, distinguishing live data from simulated inputs. Show an actual Gemini request and response, explaining its contribution beyond a rule-based route lookup. Your testing should compare route accuracy and response latency against manual selection, including stale or conflicting reports. Your feasibility plan should identify transit-data sources, report-validation methods, inference and reward costs, and an enterprise pilot partner with a measurable adoption goal.
- Claude Opus 5.5Preliminary37/100
- Strengths
- Your problem framing is concrete and local: Baku commuters at peak hours (08:00–10:00, 18:00–20:00), with a named pain point on the Nizami–28 May corridor. Your crowdsourced reporting module, paired with EcoPoints and a tangible coffee reward, is a sensible way to tackle the cold-start data problem that most crowd-density apps face. You also documented a real failure from testing (station-name typos and slow free-text chat responses) and explained how you changed the design in response.
- To improve
- Your prototype is described as frontend-only and needing no API key. That makes it unclear what Gemini actually computes; show the prompt, the inputs (crowd reports, schedules) and a real model output that differs from a simple shortest-path lookup. Replacing free text with dropdowns fixed the typos but also removed most of the AI's role, so make clear where AI is still essential. Add a few measured test routes comparing your recommendation and travel time with Google Maps or the current BakuBus/metro information. State running costs, a data source beyond crowdsourcing, and a concrete next step such as a pilot with a business or transit partner, since this track asks for an enterprise workflow and an adoption metric.
The team
Turqut İbrahimov
designer
Emin BalammədovLead
developer
Asif Süeymanlı
data