Risk reduction on AI investment — from whiteboard to production.
Francis Paul C. Flores builds LLM and automation systems that survive contact with real data, real users, and real edge cases. Fine-tuned NLP, agentic workflows, and analytical dashboards — shipped, not demoed.
Top 0.1%
of 32M+ members · Notebooks Expert
High-impact papers
The Lancet · Nature Medicine · The Lancet Public Health · JAMA Network Open
KE Excellence
Knowledge Exchange Excellence Award (2024)
95%+ Faster
Reporting turnaround reduction · 20 hrs/week reclaimed
5M+ Rows
Automated end-to-end pipelines processing government data
Production-grade AI — not demos.
Four capability areas built from real delivery history. Each maps to a specific class of business bottleneck that AI can eliminate.
Agentic Workflow Automation
Multi-agent orchestration, automated reporting pipelines, and end-to-end data workflows that reclaim hundreds of analyst-hours.
Fine-Tuned Classification Engines
Production multilingual text classifiers — XLM-RoBERTa, LoRA/QLoRA — that replace manual coding work with calibrated, auditable models matching human-level agreement in a fraction of the time.
Predictive & Forecasting Systems
Forecasting and risk-modeling pipelines — ARIMA, Prophet, LSTM, mixed-effects panel models — that convert historical and longitudinal data into forward-looking, decision-ready projections, validated to the same rigor as peer-reviewed research.
Analytics & Decision-Automation Dashboards
Python/Plotly/Shiny reporting suites that turn raw data into executive-ready visualizations with zero manual overhead.