In the high-stakes world of pharmaceutical development, where the average cost to bring a single drug to market has soared to $2.67 billion, the bottleneck is rarely a lack of ambition—it is a lack of actionable insight. As pharmaceutical giants navigate increasingly complex clinical trials, they are turning to a new breed of technology partner to bridge the gap between raw biological data and successful therapeutic outcomes.
Immunai, a pioneering startup constructing a foundational AI model of the human immune system, has solidified its role as a mission-critical partner for global pharmaceutical leader AstraZeneca. In a recent move that underscores the depth of this integration, the two companies have expanded their oncology collaboration for the third time. Under the terms of this latest agreement, Immunai is eligible to receive up to $37.5 million through 2027 as it embeds its AMICA-OS (Atlas of Multisystem Immune-Cell Analysis) platform deeper into AstraZeneca’s expansive clinical development pipeline.
A Growing Partnership: From Pandemic Roots to Global Scale
The relationship between the two entities is not merely a transactional service agreement; it is a multi-year evolution of collaborative science. While the official partnership began in late 2022, the professional history between the teams stretches back five years, originating in the volatile, high-pressure environment of the COVID-19 pandemic.
“We’ve known the AstraZeneca team for about five years,” says Noam Solomon, CEO of Immunai. “What began as a targeted effort in oncology clinical programs has since widened significantly.”
The scope of this collaboration has become a bellwether for how big pharma intends to utilize AI. In October 2025, the partnership expanded to include inflammatory bowel disease (IBD), signaling a strategic pivot toward broader immunological applications. According to Solomon, the roadmap is clear: the collaboration started in immuno-oncology, moved into general oncology, branched into immunology and inflammation, and is now encroaching on cardiovascular inflammation, neuroinflammation, and even the metabolic complexities of obesity and diabetes.
“The common thread is the immune system,” Solomon notes. “Whether it’s a cancer cell or a metabolic pathway, the immune system is the mediator. If you can model that, you can unlock insights across almost every major therapeutic area.”
The Challenge of Scale: Orchestrating Innovation
AstraZeneca is a behemoth, employing approximately 95,000 people and managing a staggering portfolio of more than 100 Phase 3 clinical trials. Coordinating with an organization of this magnitude requires more than just high-level algorithms; it requires deep operational integration.
“Over the years, there are many dozens of people on their side and dozens on ours,” says Solomon. “We don’t work with a single point of contact. We work with AI and data science teams, translational medicine departments, and clinical development leads. Each group covers different indications, and we act as the connective tissue, ensuring that data flows from the clinic to the laboratory and back into the model.”
For a startup to function as the “plumbing” for such a massive enterprise, it must solve for the logistical friction that usually stalls innovation. By standardizing the way clinical data is interpreted and analyzed, Immunai has positioned itself as an essential utility for AstraZeneca’s R&D engine.
The ‘Plumber’ Philosophy: Fixing Expensive Bottlenecks
Solomon famously describes his company’s role as that of a high-end plumber. In an industry where “data liquidity” is often hampered by silos, fragmentation, and technical debt, Immunai fixes the “expensive plumbing issues” that prevent scientists from seeing the full picture.
This plumbing work begins with data manipulation at a massive scale. When a clinical trial generates thousands of patient samples, the data is often stagnant, residing in cold storage or disparate databases. Immunai’s process transforms these biological specimens into a digital twin of the patient’s immune system. By applying proprietary immune profiling, the platform extracts clinical covariates that were previously invisible to standard analytical methods.
Pharma partners generally approach Immunai when they hit a wall that existing infrastructure cannot overcome. Common challenges include:
- Patient Stratification: Identifying which patients will respond to a specific treatment before the trial begins.
- Biomarker Discovery: Pinpointing the specific immunological triggers for toxic events.
- Combination Therapy Optimization: Determining the right agents to pair when monotherapy fails to hit efficacy targets.
- Dosing and Scheduling: Refining the clinical protocol to maximize patient safety and survival outcomes.
From Samples to Insight: The Power of Single-Cell Resolution
What distinguishes Immunai from the glut of AI-pharma startups is its "wet-lab-first" approach. While many competitors attempt to mine existing, often low-quality public data, Immunai insists on generating its own high-fidelity datasets.
“The signal already exists, but it’s hidden in the clinical patient samples sitting in your biobanks,” Solomon explains. “The starting point is always the same: send us the samples from the clinical trials. We translate those specimens into digital data using single-cell multi-omic profiling.”
The laboratory at 430 East 29th Street in New York functions as the heart of this process. Every project involves an analysis of how the immune system evolves before and after a therapeutic intervention. Immunai treats each patient profile like an “immune MRI.” This entails looking at 10,000 cells per sample, with each cell measured for roughly 37,000 gene expressions, 75 surface proteins, and VDJ sequencing.
This level of resolution is a game-changer. By correlating immune surrogate endpoints with actual clinical outcomes—such as progression-free survival—the team can pinpoint the exact immunological features that dictate whether a drug succeeds or fails.
Why Resolution Trumps Scale
The industry is obsessed with "Big Data," but Immunai argues that the market has confused volume with value. The AMICA database currently contains more than 300,000 samples, but only 50,000 are at single-cell resolution. Solomon is adamant that the latter is what truly moves the needle.
“A lot of the ‘big numbers’ in this field don’t actually lead to better decisions because the data was collected without depth,” says Solomon. He offers a compelling analogy: “If you try to scale a low-resolution, black-and-white photograph, it stays blurry. You’ll never be able to distinguish between green and blue. If your goal is to understand the nuanced colors of the immune system, you need high-resolution data. Without that, you’re stuck.”
The foundation model architecture allows Immunai to extract meaningful insights even from small cohorts, such as trials with only 20 patients. Because the model has been trained on a massive foundational dataset, each new cohort acts as a layer of reinforcement. It allows the model to “resolve the signal” where traditional statistical methods would see only noise.
Implications for the Future of Medicine
The implications of this technology extend far beyond the current collaboration with AstraZeneca. By building a comprehensive model of the human immune system, Immunai is essentially creating a blueprint for the future of precision medicine.
In April 2025, the company collaborated with the Parker Institute for Cancer Immunotherapy to assemble the largest single-cell dataset for real-world immunotherapy research, drawing from over 3,700 blood samples across 1,070 patients. Furthermore, a January 2026 partnership with Bristol Myers Squibb further cements the company’s role in guiding high-level development decisions.
As Immunai continues to expand its reach, the "plumbing" it provides is becoming the standard infrastructure for modern drug discovery. By turning the chaotic, heterogeneous data of the human immune system into a structured, predictive engine, Immunai is not just helping pharmaceutical companies save money—it is shortening the time between a biological hypothesis and a life-saving therapy.
For an industry that has long been defined by high attrition rates and costly failures, the ability to see the immune system in high definition is not just a technological advancement; it is a fundamental shift in the way humanity fights disease. As Noam Solomon’s “digital plumbing” continues to connect more nodes of the clinical landscape, the promise of truly personalized medicine moves from a distant vision to an operational reality.
About the Author
