Data Infrastructure.
AI Systems.
Built for Impact.
I'm Muhammad Usman — I build data systems, AI infrastructure, and operational tools that help teams make faster decisions. From managing 200+ resource operations to building real-time AI engines at Nobel AI, I create scalable systems that turn raw data into strategic advantage.
Who I Am
I build data systems and operational infrastructure that help teams make faster, smarter decisions. My work sits at the intersection of data analytics, AI engineering, and operations management.
At Nobel AI, I built the real-time meeting intelligence engine — designing the API architecture and writing the technical documentation the dev team now works from. Through Data with Usman, my consultancy, I built an AI-powered B2B lead generation tool using the Anthropic API for Saudi construction clients.
Before that, I managed a 200+ rider fleet operation, building the Power BI dashboards and Excel systems that stakeholders used to track KPIs and make decisions. I've also taught Python at Stanford's Code in Place program and built ML pipelines for predictive analytics.
I'm currently based in Saudi Arabia, open to remote and relocation opportunities. I'm looking for roles where I can build the systems and data infrastructure that power investment decisions and operational excellence.
What I Build
Building the operational and data infrastructure that helps teams move faster and make smarter decisions.
Data Systems & Operational Infrastructure
For teams that need structured data systems to power faster decisions across their organization.
- Deal flow tracking & CRM systems
- Portfolio performance dashboards
- Automated reporting infrastructure
- Data architecture & system design
AI-Powered Automation & Analytics
For organizations ready to leverage AI for competitive advantage and operational efficiency.
- AI-powered research & screening tools
- Predictive analytics & forecasting
- LLM integration & AI agent systems
- Automated data processing pipelines
Business Intelligence & Reporting
For leaders who need clear visibility into operations, performance, and growth metrics.
- Custom Power BI & Excel dashboards
- KPI tracking & stakeholder reporting
- Operational workflow optimization
- Cross-platform data integration
Projects
A mix of business solutions, engineering work, and AI research.
Nobel AI — Real-Time Meeting Intelligence Engine
Virtual meetings lack structured intelligence. Soft skills, participation patterns, and decision-making signals go untracked and unanalyzed.
Built the core backend engine for real-time soft skills analysis. Designed API architecture and technical documentation the dev team now works from.
200+ Rider Fleet Operations & KPI Infrastructure
A delivery fleet of 200+ riders had no centralized tracking. Route efficiency, supplier KPIs, and performance data were scattered across spreadsheets.
Built data-driven tracking system in Excel & Power BI. Optimized route efficiency by 15%+. Improved supplier KPI ratings and reduced delivery delays.
AI-Powered B2B Lead Generation Tool
Saudi construction clients needed a scalable way to identify and qualify B2B leads without manual research.
Built an AI-powered lead generation tool using the Anthropic API. Automated prospecting pipeline that delivers qualified leads at scale.
Telecom Customer Churn Prediction
Telecom company losing customers with no predictive system to identify at-risk accounts early enough to act.
Built ML model using Logistic Regression with class balancing. Delivered business insights and actionable retention recommendations from analysis.
Operations KPI Dashboard
Management had no real-time visibility into project status or budget burn — decisions were delayed by days.
Live dashboard updated daily — decision time reduced from days to minutes. Stakeholders gained full operational transparency.
Model Blind Spots Evaluation Dataset
AI models have systematic blind spots that standard benchmarks miss — no standardized way to evaluate them.
Published research-grade evaluation dataset on Hugging Face for testing model blind spots. Contributes to evaluation methodology in AI systems.
→ More projects coming as I build them.
Skills & Stack
Writing
Breakdowns, lessons, and real talk about data, AI, and building systems that work.