Silicon Valley was once a key region for the defense industry. It produced chips and other technologies that helped America overtake the Soviet Union in the Cold War. The region is now known for its defense and silicon products, but not for its consumer internet products like Netflix, Google, Facebook, and Facebook.
A few startups are trying to revive this important government-industry nexus, as more defense planners in Washington push for America's technical advantage. Vannevar Labs, one of these startups, has reached some new milestones in their quest to replace traditional defense contractors with Silicon Valley entrepreneurial acumen.
The company was just about to launch in late 2019 after raising a seed round of $4.5 million. I spoke with them again shortly thereafter. As it has developed a product and gained traction in the defense industry, the company has been quiet and head down for the past two-years. It is now ready to share some of the results of all this hard work.
In January this year, Vannevar Decrypt was launched by the company. It was designed to handle foreign language natural language processing. This involves organizing foreign data and resources collected by the intelligence community, and then translating and interpreting these documents for foreign policy decision makers. Brett Granberg, the CEO and cofounder of the product, stated that it went from one deployment into a dozen.
The company also raised $12 million Series A funding in May from Point72 and Costanoa Ventures, with General Catalyst taking part. The startups seed round was co-led by GC and Costanoa Ventures.
The company is now on a hiring spree. The company has now grown to 20 employees. Last week, Scott Sanders was hired by the firm as its head of business development. Sanders, who had been with the company for several years, was one of the first employees of Anduril. Granberg reports that Vannevar also appointed John Doyle, a Palantir veteran who headed its national security department, to its board. He says that today's team is evenly split between technologists and national security personnel, and that they are on track to double in the coming year.
Granberg believes that with a few years of experience, he has perfected what he considers to be the best model for defense technology startups to enter the tough market at the Pentagon or across Northern Virginia.
The first is to put a lot of effort into getting to the actual users and learning about their work. It is impossible to transfer what works in a large corporation to defense because the functions performed by intelligence and defense personnel are very different. Granberg referred to the Department of Defense as stating that you need both the DNA of understanding new technology, and the DNA of deep understanding many different use cases within DoD.
This directly influenced the development of Decrypt over time. We began to focus on counter-terrorism and, as the government moved away, we moved to foreign actors that were more important. After our first two deployments, it is possible to iterate very quickly.
He strongly rejects the popular belief in defense procurement circles, that dual-use technology can be used in both defense and commercial applications. He explained that some of the most critical mission problems for the government are those where they spend the most money and have the greatest interest. These are contexts in which commercial off-the shelf products (known as COTS in industry parlance), are the least useful. He said that startups focusing on defense cannot use their bandwidth to also learn about commercial uses cases.
He even predicted that there would be a lot more companies that raise a lot but fail to execute the dual-use model.
He argues that startups in defense tech should move away from the idea that every company should be on one platform and to an organizational model in which a company can offer multiple products to scale. Granberg says that each product can generate a few hundred million dollars in revenue, but it's difficult to grow a company's size if there isnt product development.
Granberg stated that Vannevar Labs is always looking for new product lines to grow. He said that Decrypt is our first product, but that 10% of our energy is spent on new product initiatives. It is possible to imagine that when we reach three-to four years later, there might be nine to ten products. Shield AI and Anduril are two of the newer companies. They have focused on product line expansion.
Granberg also mentioned the other companies. He believes that there is a mutual benefit when each startup advances in the defense sector. He said that we don't view ourselves as rivals because if one of them does well, all of us do well. Startups have lots of options to explore due to the diverse mission requirements of different agencies, and the enormous budgets available in this field. Even though they often come up against big defense contractors, there is still a lot of ground to cover.
Vannevar Labs' next goal is to make its Decrypt product a program-of-record, which would ensure it an adequate level of sales and revenue over the long term. This would not only be a major step in the company's long-term path, but it is a significant milestone.
