Biotech & AI Consulting

De-Risking Drug Development
Through AI & Deep Science

Ensemble Group Holdings combines pioneering expertise in radiogenomics, medical AI, and drug discovery to help biotech and pharmaceutical companies make smarter decisions faster.

Right Target Right Molecule Right Patient Right Test

Subscribe to Ensemble Research →

Scroll

What We Do

Our Services

Specialized consulting at the intersection of biotechnology, clinical science, and artificial intelligence.

Clinical De-Risking & Drug Intelligence

We provide rigorous, mechanism-based assessments of drug programs to help investors and pharma partners make informed decisions. Our proprietary analytical frameworks integrate structural biology, biophysics, and multi-dimensional clinical data to answer the questions: "Why and Where will this drug succeed or fail?"

  • Due diligence for biotech VCs and pharma BD teams
  • In-licensing evaluation (including China-to-West asset transfers)
  • Clinical trial design optimization
  • Biomarker discovery and validation

The Two Questions Every Drug Must Answer

Q1: Is it disease-modifying?

Does the drug exert sufficient biophysical control to drive durable disease modification? Or is it just moving a biomarker without changing outcomes?

Q2: In which patients?

Have we identified the patient ecosystem where this mechanism is a critical dependency? With a recruitable biomarker for trial enrollment?

"We identify WHERE in the causal chain assumptions are unvalidated — before Phase 3 investment."

AI Drug Design & Computational Biology

We apply advanced computational methods to accelerate therapeutic design. Our capabilities span small molecules, peptides, antibodies, and process optimization.

  • Conformational biology analysis for target validation
  • Drug molecule design and optimization
  • Mechanism-of-action characterization
  • Predictive modeling for clinical outcomes

Medical AI Development & Deployment

We have designed, validated, and deployed clinical AI systems at scale. Our experience includes multi-center international trials and production deployment across national healthcare systems.

  • Computer vision for medical imaging (radiology, pathology)
  • Multi-modal data integration (imaging + genomics + clinical)
  • Clinical validation study design
  • Large scale development, deployment, and integration of clinical AI systems

Our AI systems are currently deployed in production across an entire national healthcare system, serving millions of patients.

Radiogenomics & Biomarker Discovery

Dr. Kuo founded the field of radiogenomics — the integration of multi-scale and multi-modal data types to enable discovery of deep, contextual, non-invasive biomarkers. This pioneering work, published in Nature Biotechnology and PNAS, initially established how imaging can decode molecular signatures of disease.

  • Non-invasive biomarker development
  • Imaging-genomic integration for precision medicine
  • Oncology biomarker discovery (lung, breast, liver, kidney, prostate, brain)
  • Companion diagnostic development

How We Work

Our Approach

The Builder

Designing therapeutics that impose biophysical control over disease biology — not just target binding.

The Auditor

Clinical de-risking for your therapeutic — validating assumptions before Phase 3 investment.

TARGET ID

MOLECULE DEV

PK/PD

PRE-IND

PHASE 1

PHASE 2

PHASE 3

FILING

Where We Insert

Beyond here: capital committed

"The fusion of multi-scale data enables exponential, not linear, insight, enabling us to identify WHERE in the causal chain assumptions are unvalidated — before Phase 3 investment."

The Two Questions Every Drug Must Answer

Ensemble's Integrated Design Process

Q1 + Q2 as design drivers (not just checkpoints)

Q1: Is it disease-modifying?

Biophysical control for durable change

Drives: state selection, mechanism design, resistance prediction, durability engineering

Q2: In which patients?

Patient ecosystem where mechanism matters

Drives: context mapping, biomarker selection, trial design, stratification strategy

"Q1 + Q2 don't just validate — they drive design from Day 1"

Traditional Path Sequential: design first, ask questions later

Target ID

Literature, genetics

Screening

HTS, phage, DEL

Hit-to-Lead

Binding, selectivity

Lead Opt

Med chem, ADMET

PK/PD

In vivo

THEN ask:

Q1? Q2?

(Often too late)

Problem: Optimizes for assay metrics, discovers clinical issues downstream

Ensemble's Approach Q1 + Q2 integrated into design — not asked after

We can augment your traditional path OR provide AI-guided rational design

Augment traditional workflow

Your CRO runs screens & assays — We add Q1+Q2 intelligence at each step

  • Re-rank hits by mechanism, not just affinity
  • Guide optimization toward clinical viability
  • Predict resistance trajectories early

AI-guided rational design

Replace/compress traditional cascade — Design with Q1+Q2 from first principles

  • Model disease-relevant target states
  • Design for mechanism, not just binding
  • Build durability into molecular design

What Q1 + Q2 integrated design means technically

Dynamic state modeling

Target exists in multiple states. We model which are disease-relevant and design to those.

→ Informs Q1: mechanism

Durability prediction

Predict resistance, bypass, and escape trajectories. Design durability into molecule.

→ Informs Q1: durable modification

Patient ecosystem mapping

Map where mechanism is a critical dependency — by patient context, not average.

→ Informs Q2: which patients

These aren't separate steps — they're integrated into every design decision

Two ways to engage

Work with your CRO

  • CRO executes screens & synthesis
  • We provide Q1+Q2 intelligence layer
  • We collaborate and guide optimization together with your CROs

Direct AI-guided design

  • We lead rational design process
  • CRO validates our designs experimentally
  • Faster, fewer iterations, clinical focus

Traditional: Design for metrics → ask clinical questions later → surprises at Phase 2

Ensemble: Q1 + Q2 drive design from Day 1 → know your answer before capital is committed

This integration is what startups and CROs don't have —
the ability to design with clinical reality built in, not bolted on.

Case Studies Available

Anonymized case studies demonstrating our first principles driven integrated approach with outcomes, are available upon request.

Contact Us

Our Impact

Track Record

A proven history of delivering measurable outcomes at the intersection of science, technology, and patient care.

0

Peer-Reviewed Publications

0

Biotech Exits

0

Drug Candidates in Phase 1/2 Clinical Trials

0

Patients Served by AI

Entrepreneurial Exits

Confluence Life Sciences & Confluence Discovery Technologies

Small molecule drug discovery company and full-service therapeutics CRO

Acquired by Aclaris Therapeutics (NASDAQ: ACLR), 2017

4 drug candidates now in Phase 1/2 clinical trials

Achilles Biotherapeutics

Pre-IND immuno-oncology therapeutics company

Acquired by Lymbus Oncology, 2020

Clinical AI Deployments

Hip Fracture Detection AI

Computer vision system for emergency X-ray interpretation

Deployed across all Hong Kong Hospital Authority EDs since 2023

Serving 7+ million patients territory-wide

COVID-19 Detection AI

Deep learning for SARS-CoV-2 detection from chest X-ray

Multi-center, 4-country international validation trial

Published in European Radiology

Scientific Recognition

Alexander R. Margulis Award

RSNA's highest research honor, 2016

Gold Medal

RSNA AI Competition, 2022

First Place

Novo Nordisk / Innocentive Challenge, 2020

Gary Becker Young Investigator

Award, 2006

Independent Research

Ensemble Research

Computation-driven analysis of drug mechanisms, resistance prediction, and biotech investment implications for targeted therapeutics. Published bimonthly on Substack.

Free Subscribers

Flagship analyses of drug mechanisms, resistance prediction frameworks, and biotech investment implications — spanning kinase inhibitors, targeted biologics, and emerging therapeutic modalities.

Paid Subscribers

Full computational depth, engine-generated insights, resistance and risk landscape assessments, and early access to new analyses powered by molecular dynamics, evolutionary modeling, and structural biology.

Who We Are

Leadership

MK

Dr. Michael D. Kuo, MD

Founder & Managing Partner

Dr. Michael Kuo is a physician-scientist and entrepreneur who founded the field of radiogenomics.

He has built and sold two drug discovery companies: Confluence Life Sciences (acquired by Aclaris Therapeutics, NASDAQ: ACLR) and Achilles Biotherapeutics (acquired by Lymbus Oncology), as well as the therapeutics contract research organization (CRO) Confluence Discovery Technologies (now a sub-unit of Aclaris Therapeutics). Through these ventures, he helped developed 5 drug candidates now in, or progressing towards clinical trials and built a full-service therapeutics CRO from the ground up.

At the University of Hong Kong, Dr. Kuo founded the Medical AI Laboratory (MAIL) and led the development of clinical AI systems now fully integrated and deployed at national scale, including a hip fracture detection system serving all emergency departments across Hong Kong.

With 87+ peer-reviewed publications, an H-index of 35, and 178+ invited lectures globally, Dr. Kuo brings unparalleled scientific credibility. Combined with his entrepreneurial track record and AI deployment experience, he offers clients a rare combination of deep expertise and proven execution.

Credentials

MD, Baylor College of Medicine

Interventional Radiology Fellowship, Stanford University

Diagnostic Radiology Residency, Stanford University

Clinical Faculty, Stanford University VA Hospitals

Fmr. Associate Professor, UCLA (Radiology, Pathology, Bioengineering)

Fmr. Assistant Professor, UC San Diego

Fmr. Visiting Professor, University of Hong Kong

Let's Connect

Work With Us

Whether you're evaluating a drug program, building clinical AI, or seeking strategic advisory in computational biology, we'd like to hear from you.

[email protected]