'Developing a New Drug Is Stumbling in the Dark Without a Map. We Are Building the GPS'

When Immunai, a machine learning firm creating a ‘GPS’ for drug development, was just getting started a leading professor was skeptical about it – now he's their close consultant

Sagi Cohen
Sagi Cohen
Immunai’s system is a data-driven 'GPS' that uses AI to say in very early stages what drugs could work and utilize our immune system, thus saving years and billions
A lab in Oxford. Immunai’s system is a 'GPS' that uses AI to say in very early stages what drugs could work and utilize our immune system, thus potentially saving years and billionsCredit: PETER NICHOLLS / Reuters
Sagi Cohen
Sagi Cohen

Three years ago, a short time after he founded Immunai, Noam Solomon spoke on the phone with a famous doctor in the U.S., a professor from a big hospital who specializes in oncology. Solomon told him about Immunai and what it was trying to do. The professor told him he was skeptic about what the company was trying to do. "But after that he became our close consultant", says Solomon.

“I recently reminded him of that line he told me when we first spoke,” recalls Solomon. “After the drug we are helping to develop showed good preliminary results, he told me: ‘At least I was honest with you.’ The point is that you need people around you who will explain to you that you’re making a mistake, challenge you and also tell you that you’re an idiot. I make sure that my advisers and employees tell me the truth, because it’s the only way we can actually improve.”

Solomon founded Immunai with Luis Voloch in 2018. Within four years the company raised $295 and reached a valuation of $1.2 billion. Immunai works on mapping the body’s immune system and understanding the different interconnections within it – using AI with the goal of producing insights and data that will speed up the process of drug development and make it less expensive. This included immunotherapy treatments for cancer: Treatments that use the body’s own immune system to battle cancerous growth.

Noam Solomon founded Immunai together with Luis Voloch in 2018. Today it is valued at $1.2 billionCredit: David Bachar

“Because I didn’t come from inside the field, for the first two years I didn’t understand anything about business and biology. The result was that I asked questions all the time, and didn’t settle for ‘that’s how it’s usually done’ explanations. That’s how an open culture of healthy skepticism developed for us,” said Solomon.

AI and cancer

The story of Immunai was born from the meeting between Solomon and Voloch: The two knew each from when they were graduates of the math and computer science departments at MIT and Harvard, respectively, and their firm began to take form when they were nearing the end of their post-doctoral research.

Voloch’s grandfather had just died of cancer. His passing first inspired Solomon and Voloch to use AI, specifically so-called machine learning, to try to improve the predictions of effectiveness for cancer treatments by studying the immune system. “We managed to improve the prediction by quite a few percent points,” said Solomon. “And then I decided to make a hard pivot in my career. Without even having a presentation deck and or anything - we were just just two guys with an idea - and we set out at the beginning of 2019. We met with [venture capital firm] TLV Partners, who invested $5.5 million in us at the pre-seed stage. A few months later, the investment grew to $20 million when venture capital fund Viola Ventures joined in. Everything happened very fast.”

But the business direction that Voloch and Solomon originally banked on was very different. “Our original goal was to work with healthcare organizations and insurance companies to find the right treatment for patients. Very quickly, we realized that the American health system was not ready for it, there wasn’t enough business interest for it to happen in the next few years,” he added.

“While we worked, we began seeing that it was drug companies that actually had a demand for our data. Within a few months, we built a lab in New York that does deep genetic sequencing of the immune system, from samples sent to us by academic partners and drug companies.

A cancer research lab in the U.S. 'Past drug development was done by trial and error - what you could describe as stumbling in the dark without a flashlight or without a map'Credit: AFP

“As time went on, we saw that what our data actually knew to explain was why certain drugs don’t work. Today, we help drug companies understand how the next generation of the drugs they are developing will look and speed up their clinical trials.”

“The development process of immunotherapy treatments can take 15 years and cost $2.5 billion. Over 90 percent of the drugs that reach clinical trials do not manage to receive regulatory approval. I believe that it’s possible to change that - and trim off a significant junk of both time and money. That’s Immunai’s vision,” said Solomon.

So what exactly does Immunai’s system do? According to Solomon, “At the most basic and simple level, if in the past drug development was done by trial and error - what you could describe as stumbling in the dark without a flashlight or without a map. So we developed a GPS that helps this process happen in a manner that is data driven. Of course, it’s not as simple as entering your final destination and jumping into your car, there is still a lot of uncertainty and errors – but it’s possible to add intelligent data-based ‘directions’ and this improves the process of developing the drug.”

Solomon says that today the company has 30 to 35 projects in cooperation with drug and biotech companies, hospitals and universities. “One of our partnerships – cellular treatment that we developed with a hospital in Texas – showed promising results a few weeks ago. We have new drugs that we are helping develop, and the data speak for itself.”

The company, he says, is currently already making revenue. “But don’t ask how much. We could increase our revenue five fold if our goal was to make money now. But the goal for us now is to forge new collaborations that will give us new samples, data and intellectual property – that will allow this monster that we are building to be even more monstrous. Our real test will be when a drug that we helped speed up its development will reach clinical trials.”

Despite their initial successes, the question does arise about the firm’s high valuation - over one billion. When asked where the high valuation comes from, Solomon admits that though it’s a “good question,” he stresses he’s not the one who needs to answer it, but those who invested in the firm based on the valuation.

“I can say what the value is derived from: Immunai’s claim is that we have the ability, facilitated by AI-driven mapping of the immune system response, to measure and know if a drug works already at an early stage. This means we can accelerate clinical trials in which we see a desired response, and stop those where we see the wrong response before additional resources are poured into it. It seems that we showed investors data that support this claim. All the rest is just noise.”

Biologists and other scientists sometimes cast doubt on the ability of computing and artificial intelligence to solve complex biological problems.

Neither Solomon and Voloch are biologists, but Solomon says they brought what he describes as three of the world’s leading biologists, and helped them build a cadre of scientists from schools like Stanford and Berkeley to support their work and the vision. “They understood that the way immunological drugs are developed today is low-tech, and it’s possible to bring engineering and computational models to improve it. Immunologists, if I can make a generalization, are similar to mathematicians in that its is easy to communicate with them.”

Nonetheless, Solomon admits that there is a massive hype around AI in biology and drug development. When asked if its perhaps excessive, he says: “Absolutely yes. If I would choose the name of the company today, I wouldn’t use AI, but something different that represents it. It’s not because I don’t believe in AI, but anyone who says AI can cure cancer – is simply abusing this term. Some of the price that companies pay now in valuation is because of this irresponsibility, the sector is in many senses out of control.”

A personal roller coast

Solomon tells about his personal challenges as part of managing a startup. “For a certain period of time, when we were in one of the funding rounds, we bought two companies and I dealt with the death of a beloved relative from cancer. During her shiva I got sick with Covid, and a few months later, one of our employees died of cancer. That period changed my state of mind.”

“I can’t see myself as the CEO of a company focused only on money and valuation; I do this because I think about my relatives and the people who get sick with and died of cancer.”

The result, he says, is a sense of urgency in his management style: “I discovered that I like roller-coasters and getting phone calls at 3 in the morning when things don't work. I’m a quite demanding manager, I work with a sense of urgency. My biggest challenge is to arouse the desire in other people and solve the problems themselves, even when they look impossible.”

Immunai employs 145 people today in Tel Aviv, San Francisco, New York, Zurich and Prague. Most of them are scientists from a range of fields: Biology, medicine, immunology, engineering, computer science, machine learning, bioinformatics and others. “You need a continuous and strong culture of dealing with crises,” said Solomon. “We are building the spaceship in the middle of the flight.”

There are large wage differences between biologists and engineers. How do you deal with the tensions over the issue?

“We pay biologists more than in other places, and programmers less than in other places. It doesn’t mean there aren’t gaps, but they aren’t significant. Here we pay less than in cyber or the technology giants, and people still join us. People are looking for meaning, to know they helped people receive better medical treatment,” Solomon said.

This interview was conducted as part of the “How I built a startup” series held in partnership with the Gindi Towers TLV project in which leading entrepreneurs share their professional stories

Click the alert icon to follow topics:



Automatic approval of subscriber comments.

Subscribe today and save 40%

Already signed up? LOG IN


Crowds at Israel's Ben-Gurion International Airport, in April.

U.S. Official: West Bank Entry for Palestinian Americans Unrelated to Israeli Visa Waivers

Haaretz spoke with several people who said they had fled Ukraine, arrived in Israel,  and were asked to undergo DNA tests in order to establish paternity.

'My Jewish Grandmother Has a Number on Her Arm, Why Does Israel Greet Me This Way?'

FILE PHOTO: A Star of David hangs from a fence outside the dormant landmark Tree of Life synagogue in Pittsburgh's Squirrel Hill neighborhood in 2021.

American Judaism Is in Decline. That's Great News for American Jews

People taking part in the annual "March of the Living" to commemorate the Holocaust, between the former death camps of Auschwitz and Birkenau in Oswiecim, Poland, four years ago.

It’s Not Just the Holocaust. Israel Is Failing to Teach the History of the Jews

 A Jewish cemetery in Warsaw, Poland.

Israel and Poland Fight Over History, Truth - and Israeli Students

A collage of the Bentwich family throughout the generations.

Unique Burial Plot in Jerusalem Tells Story of Extraordinary Jewish Dynasty