Manifest Technologies: From Yale Research Spinout to Strategic Exit
April 4, 2025 | Ariane Tom, PhD
Manifest Technologies was one of the first investments I backed as Kaleida Capital was taking shape — and one of the earliest validations of our NeuroAI thesis. I first encountered the company at a formative stage in 2023 and led the investment evaluation, bringing on Flip Sabes — a longtime friend and colleague in neurotechnology —to assess the technical foundation of the platform. Roughly one year later, Manifest was acquired by Johnson & Johnson.
What made Manifest the right first investment for Kaleida — and what factors enabled the company to reach a strategic outcome so quickly?
The answer lies in the convergence of scientific depth, commercial urgency, and a technical platform designed around one of CNS drug development’s hardest problems: measuring the brain with enough precision to stratify patients and predict treatment response. Spun out of Yale research in computational neuroimaging, Manifest developed a platform to map brain structure, function, and chemistry onto patient-specific patterns of symptoms and potential drug response. For Kaleida, the company became an early proof point for precision neuroscience, a key category within our NeuroAI thesis.
From Yale Research to Precision Neuroscience Platform
Manifest Technologies did not begin as a conventional software company. It emerged from more than a decade of neuroscience research at Yale University’s School of Medicine, where co-founders Dr. Alan Anticevic and Dr. John Murray were building computational frameworks to understand how the brain organizes itself — and how that organization changes in disease.
The scientific foundation was unusually strong: computational psychiatry, neuroimaging, brain-behavior mapping, pharmacological neuroimaging, and large-scale neural data analysis were not superficial additions to the platform — but foundational elements of the product.
The core scientific question was also commercially important: can symptoms, cognition, brain structure, brain function, and molecular targets be mapped in a way that becomes useful for drug development? That question matters because CNS drug development remains one of the hardest areas in medicine. Psychiatric and neurodegenerative disorders are highly heterogeneous; two patients with the same diagnosis may have very different biology, symptoms, and treatment responses. Traditional diagnostic categories often do not map cleanly onto the underlying neural mechanisms of disease.
In 2023, Manifest Technologies was incorporated as a Yale University spinout, operating under an exclusive license for its proprietary software and patents. The company’s platform applied machine learning and computational neuroimaging to multimodal datasets, including anatomical MRI, functional MRI, PET imaging, clinical measures, behavioral data, and gene-expression maps. The commercial relevance was clear: pharmaceutical companies needed better tools to select patients, understand drug target engagement, detect early efficacy signals, and improve the probability of success in CNS clinical trials.
In Kaleida’s original diligence, this was the central insight. Manifest was not simply creating better brain images. It was building a software layer that could convert complex neural and behavioral data into patient-specific maps with potential utility across CNS R&D.
The Platform: Computational Neuroimaging for CNS Drug Development
Manifest’s core platform, NAIO™ — Neuroscience and Artificial Intelligence Optimized — was designed to integrate and analyze large-scale multimodal neuroimaging datasets. The important point is not simply that the company used AI. What made Manifest compelling was not simply that the company used AI, but the specificity of the use case: applying machine learning to multimodal brain and behavioral data to generate individualized brain-behavior profiles for patients with complex CNS disorders.
In psychosis-spectrum disorders, for example, patients are often grouped by diagnostic labels such as schizophrenia, schizoaffective disorder, or bipolar disorder with psychosis. These labels can obscure meaningful biological variation. Manifest’s approach sought to move beyond diagnosis alone and characterize patients based on underlying neural, behavioral, and molecular signatures.
This is where the technology became powerful:
Manifest’s platform could generate what I’d call a patient-specific brain fingerprint — a structured profile of brain activity, connectivity, symptom expression, and molecular relevance.
In practical terms, that fingerprint could help estimate whether a patient was more or less likely to respond to a particular therapeutic mechanism.
That matters because patient selection is one of the largest sources of failure in CNS clinical trials. If a trial enrolls patients with the same diagnosis but different underlying biology, the signal of a potentially effective drug can be diluted or lost entirely. Manifest’s platform addressed that problem through four core capabilities:
Multimodal data integration: combining MRI, fMRI, PET, clinical, behavioral, and gene-expression data into a unified analytical framework.
Brain-behavior mapping: linking brain structure and function to symptom profiles and cognitive measures to support more precise patient segmentation.
ML-enabled patient stratification: identifying patients whose neural profiles aligned with a drug’s expected mechanism of action.
Target engagement and response prediction: comparing patient-specific neural maps with pharmacological and molecular target maps to assess whether a therapeutic intervention was likely to engage relevant brain circuits.
This was the core of Manifest’s value proposition: not imaging for imaging’s sake, but computational neuroimaging as a decision-support layer for CNS drug development.
Why Manifest Fit Kaleida’s NeuroAI Thesis
At Kaleida Capital, we define NeuroAI as the convergence of neural data, artificial intelligence, and computational systems that can transform how we understand, treat, and augment the brain. Manifest fit within the precision neuroscience category of this thesis: companies that use neural data to improve diagnosis, patient segmentation, therapeutic targeting, and clinical decision-making.
This category is important because some of the most valuable NeuroAI companies may not look like conventional “neurotech” from the outside. They may not be implant companies or consumer devices. Instead, they may be infrastructure companies that determine how neural data becomes clinically and commercially useful.
Manifest was a clear example. The platform used machine learning not to replace biology, but to extract more signal from complex biological data. For pharma, it helped address several high-value questions:
Which patients should be included or excluded from a trial?
Is the drug engaging the intended neural target?
Can treatment response be detected earlier?
Can failed trials be redesigned or rescued through better patient stratification?
Can brain circuit activity be linked back to molecular targets or gene-expression patterns?
That is why I viewed Manifest not only as a CNS software company, but as part of the emerging infrastructure layer for NeuroAI-enabled drug development.
The Problem and the Scientific Breakthrough
CNS drug development has historically suffered from low approval rates, long timelines, and expensive late-stage failures. In oncology, biomarkers have transformed how trials are designed and how patients are selected. In CNS disorders, the equivalent biomarker infrastructure remains far less mature. Many trials still rely heavily on subjective endpoints, symptom scales, and broad diagnostic categories that may not reflect the relevant biology.
For disorders such as psychosis-spectrum disease, this problem is especially acute. These patient populations are deeply heterogeneous. Symptoms, cognition, neural circuitry, and molecular mechanisms may vary substantially across individuals, even within the same diagnostic category.
One of the most important pieces of Manifest’s scientific foundation was its work mapping brain-behavior relationships across psychosis-spectrum disorders. In the eLife study, researchers analyzed a cohort of 436 patients across psychosis-spectrum diagnoses and derived a lower-dimensional symptom space across psychopathology symptoms and cognitive deficits. These symptom axes mapped onto reproducible brain maps, and the study showed that a univariate brain-behavioral space model could support stable individualized prediction.
That point is important:
The goal was not simply to classify patients by diagnosis. The goal was to understand how individual variation in symptoms and cognition mapped onto individual variation in brain networks.
This is the core technical insight behind the platform: it could generate a structured brain-behavior fingerprint for each patient and relate that fingerprint to treatment mechanisms.
For psychosis-spectrum disorders, where patients are often misunderstood and poorly served by existing therapeutic frameworks, that capability is especially meaningful. It creates a path toward a more precise, biologically informed model of psychiatric disease. In the broader context of CNS drug development, that is exactly what pharma needs: better ways to match the right patients to the right drugs based on measurable biology.
What the Exit Signals
Manifest’s acquisition by Johnson & Johnson was more than a positive outcome for an early Kaleida investment —
It was a market signal.
The exit demonstrated that AI-enabled neuroimaging is not merely an academic research tool; it can become a strategic asset for pharmaceutical companies developing CNS therapeutics. It also validated several principles that remain central to Kaleida Capital:
Precision neuroscience is an investable category. Companies that use neural data to improve patient stratification, trial design, and treatment prediction can create meaningful enterprise value.
Scientific depth can become commercial advantage. The decade of research preceding Manifest’s commercialization was not academic overhead; it was the foundation of the company’s defensibility.
Domain expertise in the investment seat matters. Understanding the neuroscience made it possible to evaluate the opportunity before it was obvious to generalist investors.
NeuroAI is not limited to brain-computer interfaces. Some of the most important NeuroAI companies will be platforms that translate complex neural data into clinically actionable intelligence.
Strategic pharma demand is real. For major pharmaceutical companies, better measurement in CNS drug development is not optional; it is central to improving R&D productivity.
For Kaleida, Manifest became an early proof point for the investment strategy we continue to build: identifying technically rigorous, commercially relevant NeuroAI companies before they become consensus. The company showed that academic neuroimaging research could become a strategic platform for pharma — helping drug developers understand patients, brain circuits, and therapeutic response with greater precision.
That is why this exit matters: it was not simply a successful transaction. It was validation of a category, and of Kaleida’s ability to recognize that category early.
Ariane Tom, PhD is Founder and Managing Director of Kaleida Capital, a NeuroAI-focused venture capital platform investing across neural data, precision neuroscience, brain-computer interfaces, and frontier AI. A Stanford-trained neuroengineer and frequent speaker on neurotechnology and AI, she writes on the companies, technologies, and market shifts shaping the future of human-machine intelligence.
Kaleida Capital is focused on the emerging NeuroAI infrastructure layer — where neural data, frontier AI, and next-generation computation are beginning to reshape healthcare, intelligence, and human capability.
Connect with us if you are building, investing in, or studying the convergence of neuroscience and artificial intelligence — we welcome the conversation.