Designing a Better Future with AI-Driven Antibody Engineering

Welcome to the Genotic blog—a space where we, as founders and researchers, share the story of our journey, our aspirations, and our steadfast commitment to making the world a better place. This inaugural post marks the beginning of a series dedicated to exploring how artificial intelligence can revolutionize drug discovery, antibody development, and ultimately, patient care. We believe technology is most powerful when fueled by genuine purpose, and at Genotic, our purpose is clear: harness AI to transform the way we design antibodies—without the use of animals—while upholding the highest standards of innovation and quality.

Our Mission: A Better World Through AI

From the very beginning, Genotic was founded on one guiding principle: technology should serve to improve lives. We believe AI holds incredible potential to change how we tackle some of humanity’s toughest challenges—most importantly, how we develop life-saving treatments. Our mission is to design antibodies quickly, precisely, and ethically, aiming for a future where personalized therapies become the norm rather than the exception.

Over the past three years, we have focused on creating tools and pipelines that streamline the entire antibody engineering process. We started by modifying AlphaFold to model protein structures more accurately for our specific use cases. In fact, we tested over 50 different approaches—from custom transformer architectures and diffusion-based neural networks, to next-generation methods like AlphaFold 2 and AlphaFold 3. Through trial, error, and unwavering determination, we have developed a powerful platform to generate and verify highly selective antibodies.

A Journey of Challenges and Learning

Bringing our vision to life has been anything but simple. We encountered countless obstacles, made mistakes, and often felt like we were venturing into uncharted territory. Early iterations of our models failed to predict accurate binding sites. Some of our designed proteins wouldn’t express properly in E. coli or HEK cells, and purification processes revealed contaminants we hadn’t accounted for. Despite these setbacks, every failure was an invaluable lesson, pushing us to refine our methods, deepen our understanding of protein structures, and optimize our computational pipelines.

Those three years of intense research and development now yield results we can be proud of. We have validated over 100 distinct antibodies, each designed to bind with high specificity to a single target. Currently, our models are running on nearly 3,000 new targets, all with one overarching goal: to disrupt a status quo where many antibodies lack the specificity required for advanced therapeutic applications. By computationally verifying each candidate’s affinity and selectivity, we can ensure it binds to the intended target—and only that target—greatly reducing the risk of off-target effects.

Our Partnership with NVIDIA

We owe a significant portion of our progress to cutting-edge computational resources. A key milestone in our journey has been joining the NVIDIA Inception program as an official partner—a relationship that has greatly accelerated our research and development. We extend our deepest thanks to NVIDIA for supporting our vision, providing us access to advanced GPUs, and enabling us to push the boundaries of AI-driven protein engineering.

Harnessing Unprecedented Computational Power

At Genotic, we are proud to run one of the most robust GPU clusters dedicated to antibody design. Our infrastructure includes:

Over 200 GPUs, with 32 of them being the remarkable H100 units.
A 1.6 terabit InfiniBand network, ensuring rapid data transfer and minimal latency.
In-house servers meticulously built and optimized for high-performance AI workflows.

Over the past eighteen months, these 200 GPUs have performed more than 100,000 GPU-days of calculation (and counting). On top of our in-house resources, we’ve leveraged over $200,000 in cloud computing on Google Cloud and AWS to further speed up our experimentation. This combined computational might has made it possible to produce results that were almost unimaginable just a few years ago.

Accelerating Antibody Discovery to 48 Hours

With our current suite of AI algorithms, we can design and verify the structure of a selective antibody in as little as 48 hours of computational time. Once a lead antibody is identified, we move swiftly into production—our automated and highly efficient laboratory allows us to go from digital design to physical samples in under 20 days. For us, this is more than a technical achievement; it is a leap forward that could reshape how we tackle acute and chronic diseases alike, especially in scenarios that demand rapid deployment of new therapeutic agents.

State-of-the-Art Laboratory

Our commitment to excellence extends far beyond computer models. We have built a fully integrated laboratory in Europe, equipped with:

6 chromatographs for high-precision purification.
20 simultaneous cultures that can be initiated any given day.
Robust processes for expression in E. coli, HEK, and other systems.
Sophisticated lyophilization workflows to ensure our products have endotoxin levels below 0.5 EU and meet stringent purity standards.

To date, our facility has already produced a broad range of antibodies, enzymes, and target proteins for R&D purposes. Our small but dedicated 20-person team is continually optimizing this pipeline to shorten lead times and offer unmatched quality. While our scale may be modest, our ambitions are anything but.

Looking Toward the Future: Oncology and Beyond

At the heart of Genotic lies a passion for using scientific advancement to serve humanity. Our current focus extends beyond merely supplying research-grade antibodies; we are actively preparing to launch oncology therapies that leverage our AI-driven designs. By aiming to create treatments that are tailored to individual patients, we hope to transform the traditionally long and expensive process of drug development into something far more agile, personalized, and cost-effective.

In a world where many diseases still lack effective treatments, or where existing drugs come with significant side effects, AI-driven antibody engineering could be the breakthrough we need. We envision a future in which therapies are developed on-demand, offering hope to patients who otherwise have limited options.

Our Commitment and Invitation

Genotic stands at the intersection of AI and life sciences, guided by the belief that technology’s highest calling is to serve humanity. We have come a long way from our early attempts at customizing AlphaFold models, and each success pushes us to dream even bigger. We invite you—researchers, investors, patients, and innovators—to join us on this journey, whether by collaborating on new targets, sharing insights, or simply following our story as we forge ahead.

In subsequent blog posts, we will delve deeper into the science behind our antibody design process, the specifics of our AI models, and the breakthroughs we’ve achieved in selective binding. We’ll also share lessons learned along the way—the highs, the lows, and everything in between.

Thank you for taking the time to learn about Genotic. We look forward to shaping the future of antibody engineering together—one target at a time.

With gratitude and determination,

The Genotic Founding Team Where AI meets the drive to make the world a better place.

See you soon.

Greg Warzecha