Official profile

Omar Costilla Reyes, PhD

Building interpretable AI for health and scientific discovery.

Current focus

Interpretable AIMultimodal health dataScientific reasoningResponsible product design
Official websiteLinkedIn
Overview

From scientific research to engineered health AI solutions

Dr. Omar Costilla Reyes works across artificial intelligence, neuroscience, medical engineering, digital health, and scientific discovery. His focus is building systems that are technically rigorous, understandable, useful, and grounded in evidence.

His research background across MIT Brain and Cognitive Sciences, the Institute for Medical Engineering and Sciences, and the Computer Science and Artificial Intelligence Laboratory connects brain research, translational medicine, and computer science. That foundation informs his work on interpretable AI, neurosymbolic programming, digital phenotyping, and multimodal health data.

Through Equ Healthcare, this trajectory moves from research into product-building: privacy-conscious AI systems that organize wearable, laboratory, lifestyle, and behavioral data into guidance people and organizations can inspect, evaluate, and act on.

Profile Foundation

Current roles, research appointments, education, and core research areas

Current work focuses on translating scientific research and engineering methods into interpretable AI systems for health intelligence, multimodal data integration, and responsible deployment.

Equ Healthcare logo
Equ Healthcare

Explainable AI for Interpretable Health Intelligence

CEO & Founder

Leads Equ Healthcare, a health AI company building explainable systems that integrate wearable, laboratory, lifestyle, and behavioral data into interpretable health intelligence.

Equ AI Lab logo
Equ AI Lab

Boston, Massachusetts, USA

Director of Research

Directs the Equ Healthcare research group focused on neurosymbolic programming, causal modeling, multimodal health data, and privacy-conscious AI systems.

Research appointments formed the scientific foundation for Omar's current work as a scientist-engineer founder, connecting sensing, computational neuroscience, digital phenotyping, medical engineering, and neurosymbolic AI.

MIT CSAIL logo

Computer Science and Artificial Intelligence Laboratory (CSAIL)

Massachusetts Institute of Technology (MIT), Cambridge, USA

Research Scientist & Research Project Manager

At MIT CSAIL, worked on neurosymbolic programming for scientific discovery, developing methods that connect data-driven learning with symbolic representations, executable models, and interpretable reasoning.

MIT Institute for Medical Engineering and Sciences logo

Institute for Medical Engineering and Sciences (IMES)

Massachusetts Institute of Technology (MIT), Boston, USA

Postdoctoral research fellow, Edelman Lab

At MIT IMES, worked in a translational medical engineering environment on digital phenotyping, behavioral signals, and AI methods for mental health research.

MIT Brain and Cognitive Sciences logo

Brain and Cognitive Sciences (BCS)

Massachusetts Institute of Technology (MIT), Boston, USA

Postdoctoral associate and fellow

Picower Institute of Learning and Memory Miller Laboratory

At MIT Brain and Cognitive Sciences and the Picower Institute, applied machine learning to neural signals, learning, memory, and cortical dynamics.

The University of Manchester logo

University of Manchester

Manchester, UK

PhD Candidate & Research Assistant

Developed machine learning methods for spatio-temporal pattern recognition from sensor data, with applications in gait analysis, healthcare, and security.

University of North Texas logo

University of North Texas

Denton, TX, USA

Graduate Research Assistant

Researched sensing, localization, wireless systems, and robotics, building an engineering foundation for later work in health data and intelligent systems.

Academic training in electronics engineering, electrical engineering, sensing systems, machine learning, and computational methods.

The University of Manchester logo

PhD, Electrical and Electronics Engineering

University of Manchester

Research on spatio-temporal pattern recognition from sensor data for gait analysis, healthcare, and security.

University of North Texas logo

MSc, Electrical Engineering

University of North Texas

Research on wireless sensing, localization, robotics, dynamic WiFi fingerprinting, and machine learning.

Autonomous University of the State of Mexico logo

BSc with Honors, Electronics Engineering

Autonomous University of the State of Mexico

Foundation in electronics, systems engineering, applied computing, and technical leadership.

Selected Technical Training

Additional Technical Training

Additional training in machine learning, artificial intelligence for robotics, parallel programming, big data analytics, signal processing, wireless networks, systems modeling, and bioinformatics.

Parallel Programming - NVIDIA
Artificial Intelligence for Robotics - Stanford University
Machine Learning - Caltech
Summer School on Big Data Analytics - Caltech JPL NASA

Graduate Coursework

System Modeling and Simulation (EENG 5320)
Digital Signal Processing (EENG 5610)
Image and Video Communications (EENG 5850)
Digital System Design and Testing (EENG 5520)
Parallel Processing (CSCE 5160)
Wireless Networks and Protocols (CSCE 5520)
Bioinformatics (CSCE 6810)

Core areas across digital mental health AI, symbolic regression, neurosymbolic programming, sensing, computational neuroscience, and responsible product translation.

Digital Mental Health AI

Models for smartphone-delivered interventions, treatment response, digital phenotyping, and behavioral health signals.

Neurosymbolic Program Synthesis

Hybrid neural-symbolic systems for program synthesis, reusable abstractions, and inspectable reasoning.

Symbolic Regression

Equation synthesis methods for compact scientific models, learned concept libraries, and recursive symbolic search.

Scientific Discovery

AI methods that turn experimental, biomedical, and behavioral data into hypotheses and models people can inspect.

Computational Neuroscience

Machine learning for neural signals, oscillations, cognition, learning, memory, and cortical dynamics.

Multimodal Health Data

Wearable, laboratory, lifestyle, and behavioral data pipelines for interpretable health intelligence.

Sensor-Based Pattern Recognition

Spatio-temporal modeling for gait, footstep recognition, tomography sensors, and indoor positioning.

Responsible Product Translation

Privacy-conscious, human-reviewable systems that move research into health and scientific workflows.

Builder Trajectory

Scientific trajectory, technical methods, and product-facing focus

A path from engineering and sensing to computational neuroscience, digital mental health, neurosymbolic scientific discovery, and health AI products.

Scientific and Entrepreneurial Trajectory

Equ Healthcare logo

CEO and Founder

Equ Healthcare

Builds health AI systems that connect wearable, laboratory, lifestyle, and behavioral data with interpretable product workflows.

Health AIDigital Mental HealthPrivacy-First DesignMultimodal Health DataMetabolic HealthPersonalized MedicineOn-Device Intelligence
MIT CSAIL logo

Worked on neurosymbolic programming, symbolic regression, learned abstractions, and equation synthesis for scientific discovery.

Neurosymbolic ProgrammingSymbolic RegressionScientific DiscoveryProgram SynthesisTheory GenerationInterdisciplinary Research
MIT Institute for Medical Engineering and Sciences logo

Conducted research on digital phenotyping, smartphone-delivered interventions, behavioral signals, and AI methods for mental health.

Digital PhenotypingMental HealthMobile SensingBehavioral AnalysisTreatment ResponseNetwork Analysis
MIT Brain and Cognitive Sciences logo

Postdoctoral Associate and Fellow

Department of Brain and Cognitive Sciences, MIT

Applied machine learning to neural oscillations, learning, memory, and cortical mechanisms at the Picower Institute.

NeuroscienceMachine LearningNeural NetworksElectrophysiological Data AnalysisNeural OscillationsCortical AnalysisCognitive Modeling
The University of Manchester logo

PhD in Electrical and Electronics Engineering

University of Manchester, UK

Developed spatio-temporal pattern recognition methods for sensor-based gait analysis in healthcare and security. Thesis: Pattern recognition from raw spatio-temporal data for gait analysis in healthcare and security.

Deep LearningGait AnalysisPattern RecognitionSensor SystemsSpatio-Temporal AnalysisDeep Residual NetworksHealthcare ApplicationsBiometric Security
University of North Texas logo

Master of Science in Electrical Engineering

University of North Texas, USA

Researched wireless sensing, indoor localization, robotics, and machine learning.

Indoor PositioningWiFi FingerprintingSignal ProcessingWireless NetworksMachine LearningLocalization Systems
Autonomous University of the State of Mexico logo

Bachelor of Science in Electronics Engineering

Autonomous University of the State of Mexico, Mexico

Built an engineering foundation in electronics, systems thinking, applied computing, and technical leadership.

ElectronicsEngineering FundamentalsStudent LeadershipIEEE ActivitiesTechnical Communication

Explore My Work

Technical methods spanning health outcome modeling, digital phenotyping, symbolic regression, program synthesis, neural signals, sensing, localization, robotics, and product evaluation.

Methods spanning clinical outcome modeling, symbolic AI, mobile sensing, neuroscience, multimodal health data, spatial systems, robotics, and product evaluation.

Active method family

Health Outcome Modeling

Predictive and interpretable models for depression, anxiety comorbidity, body dysmorphic disorder, and smartphone-delivered interventions.

Profile evidence

Recent Journal of Affective Disorders and arXiv work on smartphone-delivered mental health interventions.

Specific methods
Treatment-response predictionComorbidity-aware outcome modelingCounterfactual explanationsFeature attribution review
Data and domains
MDD improvementBody dysmorphic disorderDigital CBTClinical outcome data
Typical outputs
  • Risk stratification
  • Reviewable predictors
  • Patient subgroup signals

Where AI research, engineering practice, medical science, and product development come together.

Health Intelligence from Multimodal Data

Wearable, laboratory, lifestyle, clinical, and behavioral signals connected into interpretable health models.

Digital Mental Health and Behavioral Signals

Smartphone-delivered interventions, treatment-response prediction, and mobile-sensing studies in mental health.

Symbolic Scientific Modeling

Symbolic regression, recursive equation synthesis, and learned concept libraries for inspectable models.

Neural and Cognitive Signal Analysis

Machine learning for electrophysiology, oscillations, cognition, learning, memory, and cortical dynamics.

Human-Reviewable Health AI

Models, interfaces, and documentation that support clinical and product review.

Research-to-Product Leadership

Turning peer-reviewed methods, engineering experience, and health data products into practical systems.

Selected Work

Projects and publications across health AI, symbolic modeling, sensing, and scientific discovery

Selected projects across health AI, symbolic regression, digital phenotyping, scientific discovery, sensing, AI education, and robotics.

Healthcare AI

Equ Healthcare

Health AI product work connecting wearable, laboratory, lifestyle, and behavioral data into interpretable guidance.

Healthcare AI

Digital Mental Health Outcome Modeling

Interpretable machine learning work on smartphone-delivered interventions, depression improvement, and treatment-response signals.

Scientific AI

Recursive Symbolic Regression

AAAI 2025 work on equation synthesis methods that reuse partial symbolic expressions to improve scientific model discovery.

Scientific AI

Learned Concept Libraries for Symbolic Regression

NeurIPS 2024 work on using reusable concept libraries to guide symbolic regression for interpretable scientific discovery.

Scientific AI

Neurosymbolic Programming for Science

MIT CSAIL research on AI systems that generate interpretable scientific models from experimental data.

Healthcare AI

Digital Phenotyping Platform

Mobile sensing and machine learning framework for behavioral health research, privacy-conscious phenotyping, and schizophrenia modeling.

View Related Publications

Citation-ranked publications across mental health AI, neurosymbolic systems, symbolic regression, sensing, computational neuroscience, and scientific discovery.

519

Citations

11

h-index

12

i10-index

30

Papers

Google Scholar snapshot updated Mar 17, 2026. Since 2020: 352 citations, h-index 9, i10-index 9. Metadata reviewed May 19, 2026.

Selected publications

Citation-ranked topic leaders

Showing top 5 of 30 ranked publications

#3Advances in Neural Information Processing Systems• 2024• 65 citations

Symbolic regression with a learned concept library

Arya Grayeli, Atharva Sehgal, Omar Costilla-Reyes, Miles Cranmer, Swarat Chaudhuri

Learned concept libraries for reusable symbolic regression.

Neurosymbolic AISymbolic RegressionScientific Discovery

Explore the full publication record

View citations, publication metadata, and updates on Google Scholar.

Google Scholar Profile

Advisory Services

Founder and technical advisory for health AI, digital data products, and research translation

Advisory and technical services for organizations building explainable AI, digital health systems, multimodal data products, and responsible AI strategies.

Explainable AI Strategy

Guidance for teams building AI systems that require transparent reasoning, evidence pathways, and human-reviewable outputs.

Technical Capabilities
Transparent reasoning and evidence pathway design
Interpretable model selection for health, research, and decision support
Human-reviewable AI output design
Explanation interfaces for technical and non-technical stakeholders

Build AI systems that stand up to review

Technical strategy, explainable system design, and research-to-product decisions grounded in evidence and constraints.

  • Clarify strategy before investing in the wrong data or architecture.
  • Design explainable systems for health, science, and review-heavy workflows.
  • Bridge research and product with implementation decisions that can be defended.
AI Consulting

Research depth with implementation and governance focus

Book a Consultation

Leadership

AI education, responsible adoption, research collaboration, entrepreneurship, and ecosystem work

Work across AI education, responsible adoption, research collaboration, entrepreneurship, and international technology ecosystems.

professional2025 - Present

Director of Research

Equ AI Lab

Directing research on explainable AI, multimodal health data fusion, neurosymbolic systems, and privacy-conscious AI systems.

professional2024 - Present

Chief Executive Officer

Equ Healthcare

Leading a health AI company focused on building explainable systems that integrate wearable, laboratory, lifestyle, and behavioral data.

Visit website
international2023 - Present

Member

Inter-American Committee on Science and Technology Organization of American States

Contributing to regional science, technology, and AI policy discussions across the Americas.

Visit website
community2022 - Present

Mentorship Program Coordinator

MIT CSAIL Diversity Initiatives

Coordinating mentorship to support broader participation in AI research.

academic2021 - 2025

Research Scientist

MIT Computer Science and Artificial Intelligence Laboratory (CSAIL)

Advanced computer-aided programming and neurosymbolic methods for scientific discovery.

Visit website
academic2021 - 2025

Research Project Manager

MIT Computer Science and Artificial Intelligence Laboratory (CSAIL)

Managed interdisciplinary research for the NSF-supported Understanding the World Through Code project.

Visit website

Connect

Ways to start a focused conversation

Reach out for research collaboration, technical advisory, consulting, or speaking opportunities.

Inquiries

Research, consulting, speaking, partnerships, or media.

Location

Boston, MA

Network

LinkedIn

Use the form for research collaboration, consulting, speaking, partnerships, media, or technical review requests.

Profile Sections
All sections expanded