Government Innovation Platform
InnovaFest 2026
A full-stack platform for Mexico's national innovation award (Premio a la Innovacion) with AI-powered eligibility screening and multi-venue submission management.
Industria
Government
Línea de Tiempo
14 weeks
El Desafío
Mexico's Secretaria de Economia needed a platform to manage innovation award submissions across 5 regional venues, screen applications for eligibility, and coordinate human evaluators — all within a tight timeline and with strict security requirements.
Nuestro Enfoque
We built a multi-agent AI pipeline using LangGraph with 4 specialized agents (parser, gate evaluator, cross-reference, ficha generator) that pre-screen applications through 5 eligibility gates. We used Claude Batch API for 50% cost savings on AI processing, and implemented ClamAV for document antivirus scanning.
La Solución
InnovaFest manages the full lifecycle of innovation award submissions: application ingestion with folio generation, file validation (PDF/DOC/video), AI-powered pre-filtering (5 pass/fail gates + 5 informational flags), human evaluator scorecards (15 subcriteria, weighted 0-100), venue-based role permissions, and an appeals workflow.
Los Resultados
5
Regional venues
4
AI screening agents
50%
AI cost savings
Stack Tecnológico
"The platform handled submissions from all 5 venues seamlessly. The AI pre-screening saved our evaluators significant time."
-- Program Coordinator, Secretaria de Economia
Casos de estudio relacionados
HealthTech
AI-Powered HealthTech Platform
A HIPAA-compliant AI platform with RAG-based customer support, radiologist dashboards, and an online booking system processing 10K+ appointments.
Ver caso de estudio →
LegalTech / FinTech
AI Contract Intelligence Platform
An AI-driven change control system for contract metadata analysis, serving banking, finance, and legal clients with role-based access.
Ver caso de estudio →
Education
Education Sector Microservices
A microservices architecture powering US education products with Python, Go, and React on AWS infrastructure.
Ver caso de estudio →