
Problem
Siemens innovation team receives hundreds of internal and external project proposals monthly. Manual review is time-consuming, inconsistent, and creates bottlenecks in identifying high-potential innovations worth pursuing.
Solution
Designed AI classification model that automatically sorts proposals by quality, strategic alignment, and technical feasibility. System provides standardized evaluation framework enabling faster decision-making on which innovations to fund or develop.
My Role
Product Designer & AI Solution Architect
Tech Stack
- AI classification models
- NLP
- Workflow automation
Overview
Consulting project for Siemens designing an AI-powered system to streamline their innovation proposal evaluation process. The innovation wing receives proposals from both internal employees and external partners, requiring systematic evaluation against strategic priorities and technical feasibility criteria.
Solution Design
Classification Framework
- Quality scoring based on proposal completeness and clarity
- Strategic alignment assessment against Siemens innovation priorities
- Technical feasibility evaluation using domain-specific criteria
- Resource requirement estimation for implementation
- Risk assessment for regulatory and technical challenges
AI Model Architecture
- NLP processing for proposal text analysis
- Multi-criteria classification engine
- Scoring algorithm combining automated analysis with human oversight
- Priority ranking system for resource allocation
- Integration design with existing proposal management workflows
Key Features Designed
- Automated initial screening reducing manual review time
- Standardized evaluation criteria ensuring consistency
- Priority queuing for high-potential proposals
- Dashboard for innovation team tracking proposal pipeline
- Feedback mechanism improving model accuracy over time
Value Proposition
Reduced time from proposal submission to decision from weeks to days. Standardized evaluation framework ensuring consistent assessment criteria. Freed innovation team to focus on high-value activities like stakeholder engagement and strategic planning rather than administrative review.
My Contributions
- Designed classification criteria and scoring methodology
- Developed AI model architecture and workflow integration approach
- Created evaluation framework balancing automation with human judgment
- Proposed implementation strategy for Siemens environment
Related Projects
3D Portfolio Simulator with AI Interview Assistant
Revolutionary 3D recruitment experience combining immersive WebGL graphics with advanced AI personas. Transforms traditional portfolio browsing into interactive conversations, achieving 85% recruiter engagement and 3x longer session times.

Reddit Racism Analyzer: AI-Powered Content Moderation
🏆 LSE Code Camp 2025 Overall Winner. Revolutionary AI content moderation system achieving 85%+ accuracy in racism detection. Ensemble of 3 specialized models analyzing 100+ posts in <60 seconds, reducing false positives by 40% through context-aware analysis.