Built for Environments That Can't Afford to Get It Wrong

PredictIQ AI was founded on a conviction: in regulated environments, the gap between AI ambition and production AI that actually holds up to scrutiny is a solvable engineering problem.

Why PredictIQ AI Exists

Building AI in healthcare and regulated environments is a fundamentally different engineering problem. The failure mode is not a bad user experience — it is an audit finding, a wrong medical determination, or a regulatory event. Generic AI architecture, built for environments where errors are recoverable, does not transfer cleanly.

The patterns that work — deterministic regulatory logic, evidence-grounded retrieval, bounded LLM inference, human-in-the-loop design, golden cohort evaluation — are not widely understood outside the small set of practitioners who have built production systems in these environments and watched what happens when those patterns are absent.

PredictIQ AI was founded to apply those patterns — distilled from active production work under FDA contracts, EU MDR regulatory constraints, and 38 CFR compliance requirements — as an embedded engineering partner for organizations that need them.

Founded in Naples, Florida
Serving U.S. and international clients
Specialized in healthcare & regulated AI
Active FDA contract engagement
Embedded, full-lifecycle engagements

Jorge Rivero

AI Architect & Enterprise Machine Learning Leader

Jorge is currently embedded as the sole AI engineer at ECRI Institute — the world's leading nonprofit medical device evaluation authority — building five interconnected AI pipelines for medical device horizon scanning, regulatory classification, attribute validation, EMBASE novelty mining, and FDA post-market surveillance, under an active FDA BAA contract.

As Senior Lead Data Scientist at Trajector, he designs and implements production AI systems for veteran disability claims — medical evidence retrieval, source-attributed nexus reasoning, and full implementation of 38 CFR Parts 3 and 4 regulatory logic as deterministic, auditable guardrails.

Before founding PredictIQ AI, Jorge spent eight years at Oracle Corporation building production AI for the Oracle OPERA Cloud hospitality platform — demand forecasting, room assignment optimization, federated learning for upsell prediction, and causal pricing systems. He is a named inventor on six US patents (assigned to Oracle). His academic research — zero-shot learning, metric learning, network intrusion detection — has been published at ICONIP and IJCNN with 149+ citations. An Erasmus Mundus scholar at the Universidade de Coimbra, he holds a PhD in Machine Learning and AI.

Connect on LinkedIn
Active
FDA Contract · Post-Market AI Surveillance
ECRI
Medical Device Evaluation Authority · Client
38 CFR
Regulatory Logic Implemented in Production
6
US Patents · Enterprise AI (Oracle)
149+
Academic Citations
PhD
Machine Learning & Artificial Intelligence
Erasmus
Mundus Scholar · Universidade de Coimbra
ICONIP
International AI Publication

Active Engagements

Production AI systems currently running in two regulated healthcare environments — medical device intelligence for a global health research authority, and veteran disability AI for a national VA claims company.

Active · OngoingMedical Device Intelligence · ECRI Institute

Medical Device Intelligence Platform

Five production pipelines — Horizon Scanning (72+ sources), Time-to-Market prediction, Attribute Validation, EMBASE Novelty Mining, and FDA Post-Market AI Surveillance. Grounded evidence, EMDN taxonomy coding, human-in-the-loop analyst review, active FDA BAA contract. Weekly shipping cadence.

Read the full case study →
Active · OngoingMedical Evidence AI · Veteran Disability Claims

Veteran Disability AI Systems — Trajector

Senior Lead Data Scientist — production AI systems for veteran disability claims. Medical evidence retrieval with domain-specific embeddings, source-attributed nexus reasoning, and full 38 CFR Parts 3 & 4 regulatory logic as deterministic, auditable guardrails. Golden cohort regression testing, 253+ tests across the platform.

6 US Patents

Named inventor on six US patents assigned to Oracle Corporation from Jorge's prior career — covering AI systems for hospitality, forecasting, and anomaly detection.

AI-Based Hotel Room Assignment Optimization

Optimization system for hotel room assignment decisions using machine learning.

AI-Based Hotel Demand Modeling

Demand forecasting model for hotel occupancy prediction under market uncertainty.

Federated Learning for Hotel Upsell

Hierarchical federated learning enabling portfolio-wide ML with local specificity.

Root Cause Anomaly Detection in Time Series

Graph neural network approach for identifying root causes in time-series anomalies.

Integrated Analytics for Bot Performance

Monitoring and optimization system for conversational AI performance.

Blockchain-Based Room Inventory Management

Distributed inventory management system for hotel operations.

Expertise

Large Language ModelsRetrieval-Augmented GenerationFederated LearningDemand ForecastingAnomaly DetectionZero-Shot LearningGraph Neural NetworksML OptimizationAI ArchitectureMLOpsNLPTime Series

Let's build something that ships

Start with a free AI opportunity assessment or book a strategy call.