A small lab on Tower Place is teaching machines to read the sugar coating on proteins - and using that to catch cancer that genomics misses.
It is a Tuesday morning in a gynecology clinic. A woman in her fifties is sitting on the paper, a pelvic mass on her ultrasound, two children in a waiting room, and a number on a tablet: 86. That is the accuracy with which a tube of her blood, shipped to a lab in South San Francisco, can tell her oncologist whether the mass is benign or something worse. Until 2025, that number did not exist. The blood test that gives it is called GlycoKnow Ovarian. The company behind it is InterVenn Biosciences, forty-four people on Tower Place, and what they are doing is not exactly a diagnostic. It is a new layer of biology read at clinical speed.
Genomics reads the blueprint. Proteomics counts the bricks. InterVenn reads the frosting - the sugary chains called glycans that coat almost every protein on almost every cell - and asks what flavors a tumor leaves behind. The answer, increasingly, is "more than the blueprint will tell you."
In 2016, Aldo Carrascoso - a serial-founder whose previous companies traded in payments and media, not pipettes - watched a relative get diagnosed with cancer. He sat in on a research trial, hoping genomic sequencing would explain why his family kept getting sick. It didn't. And when he asked how long it would take to analyze the sample, the answer was a year.
He met chemist Carolyn Bertozzi in the same room. She was not yet a Nobel Laureate. She would be, in 2022, for inventing the chemistry that lets scientists tag and read sugars on living cells. Carrascoso and Bertozzi, along with UC Davis chemist Carlito Lebrilla, decided the year-long answer was unacceptable. Twelve months later they had a company. Eight years later they have a commercial test.
Carrascoso has since handed CEO duties along, but the founding story is doing a lot of work in the InterVenn pitch deck: the deepest expert in the field, a frustrated operator, and a problem that genomics had quietly given up on.
Filipino serial entrepreneur. Founded Veem, Jukin Media, Glogster before pivoting to biotech in 2017 after a family cancer diagnosis.
Stanford chemist. Won the 2022 Nobel Prize in Chemistry for bioorthogonal click chemistry - the toolkit that makes glycan reading possible.
UC Davis Distinguished Professor of chemistry. One of the founding figures of modern glycoproteomic mass spectrometry.
The discovery engine. Liquid chromatography feeds samples into mass spectrometers; machine learning sifts through millions of glycan signals and finds the patterns that correlate with disease. It is the only platform in the world running glycoproteomics at clinical throughput.
The first commercial product, launched October 2025 with its own AMA PLA code. A blood draw separates ovarian cancer from benign pelvic masses with roughly 86 percent accuracy - better, in head-to-head data, than the long-standing CA-125 marker.
A liquid biopsy that predicts whether a patient will respond to immune checkpoint inhibitors - the costliest and most toxic class of cancer drugs. Tested first in melanoma, with other tumor types in the pipeline.
Quietly, multiple pharmaceutical sponsors run their biomarker discovery on GlycoVision. It pays bills, validates the platform, and gives InterVenn a second business model that does not depend on insurance reimbursement.
Series C led by SoftBank Group with Heritage Provider Network, Irving Investors, Highside Capital, Amplify Partners, Anzu Partners, Genoa Ventures, True Ventures.
Aldo Carrascoso, Carolyn Bertozzi, and Carlito Lebrilla incorporate in California.
Led by Genoa Ventures, with True Ventures, Amplify Partners, Boost VC, Prado SV.
Anzu Partners-led round to push the platform toward clinic.
SoftBank writes the check. Heritage Provider Network and Irving Investors join.
A free, permanent marketing event for the underlying science.
GlycoKnow Ovarian becomes a billable, reimbursable lab test in the United States.
InterVenn's first product on shelves, aligned with the code's effective date.
Distribution and workflow integration to push GlycoKnow into more clinics, faster.
Two thirds of the proteins in your blood are glycosylated. The sugar chains are not decoration - they decide whether a cell is recognized as "you," whether an immune cell attacks it, whether a drug binds. Cancer messes with the sugars before it makes itself visible in many other ways. For most of the last forty years, this layer of biology was studied by a small academic field with instruments that took weeks per sample.
InterVenn's wager was that you could industrialize it - shrink the workflow from weeks to hours, throw a real machine-learning stack on top, and ship the output as a clinical test a regular oncologist could order on a Tuesday. The $201M Series C was the bet that they were right. The PLA code in 2025 was the first proof.
Order GlycoKnow Ovarian when a patient presents with an indeterminate pelvic mass. Triage with more confidence than CA-125 alone.
Run your trial cohort through GlycoVision to find glycoprotein biomarkers your proteomics platform missed - and stratify responders before Phase II.
Tap a clinically validated platform without standing up your own LC-MS pipeline. Co-publish on glycoproteomic biomarkers across disease areas.
If your physician is weighing surgery for a pelvic mass, ask whether a GlycoKnow result would change the next decision. It often does.
Back in that Tuesday clinic, the number on the tablet is doing something quieter than it looks. It is not curing the woman on the paper. It is not even, by itself, a diagnosis. What it is doing is changing the next conversation - the one between surgeon and patient, between oncologist and insurance, between fear and a treatment plan. Before October 2025 that conversation was shaped by a forty-year-old marker called CA-125 and a coin-flip in a lot of cases. After October 2025 it is shaped by the sugar coatings on her proteins, read by a mass spectrometer in South San Francisco and scored by a model trained on tens of thousands of samples.
Forty-four people built that. They did not invent the biology - Bertozzi and Lebrilla had been at it for decades. They industrialized it. They got a code. They shipped a product. And they made a Tuesday in a gynecology clinic measurably less terrifying than it was the year before.
That is the InterVenn pitch in one sentence: a layer of biology, finally legible, at the speed of care.