Today, we are delighted to take our longstanding investment theme of Intelligent Applications one step further and announce the inaugural Intelligent Applications Top 40. #IA40 www.ia40.com. Co-sponsored with Goldman Sachs and supported by 40 top tier venture firms that nominated and voted on the companies they believe will change the future of software.
Investing in Intelligent Applications
Software is the foundation of modern business processes. It acts as a system of record and a system of automation – that both stores information in databases, and structures core workflows off of that information. In order to achieve these goals, software and databases need to run predictably. We believe that the next-generation of applications will leverage diverse data sources and machine intelligence to turn software in to “dataware” becoming predictive systems of intelligence.
So, what are these systems? Intelligent applications embed machine learning models and take advantage of both historical and real-time data to deliver recommendations, predictions and inferences that are either automatically leveraged by machines or made actionable by a “human in the loop.” The data around us is increasingly digital, accessible and robust and becomes the fuel to create the high quality data models that power these intelligent applications. These learning systems solve a business problem in a contextually relevant way – better than ever before. And, users find that these applications deliver rich information and insights that increase decision-making efficiency or enable superior business decisions.
Intelligent applications embed machine learning models and take advantage of both historical and real-time data to deliver recommendations, predictions and inferences that are either automatically leveraged by machines or made actionable by a “human in the loop.”
These applications also require a system of horizontal enablers – that are just starting to reach scale. Snowflake, for example, was arguably the first, modern intelligent application enabler to go public. And, we expect in 2022 there will be more.
Why another list?
Machine learning and artificial intelligence have already transformed the consumer world with the role they play in powering the “big tech” companies and they are increasingly powering the modernization of commercial technologies. As ML/AI models have matured and become more widely adopted in the developer community, their potential to be deployed in business applications has become more apparent. As early-stage investors, we have been partnering with intelligent application companies for years and watching their incredible impact on areas as diverse as finance, sales automation and life sciences. These applications will define the future of software and the next generation of computing, and we believe they deserve to be known and recognized.
Our view is that intelligent applications are largely going to supersede their SaaS application counterparts as the dominant business services of the coming decade. In fact, from a venture capital perspective, they already have.
According to Pitchbook, in Q3 of 2021 alone, venture capital investment in AI and ML topped $25.9B globally, across a record 1,395 deals.
While not all of these companies would today be considered intelligent applications, a rapidly increasing percentage of them are. There is increasing M&A activity in the field as well, as incumbents look to sharpen and modernize their capabilities, across 127 exits totaling more than $25B in disclosed value.
In addition, for those in search of a tool that will enable predictive intelligence to help inform better business decisions, start with the enabler list below.
Madrona has been investing in intelligent applications for over a decade now, beginning with companies like Placed in 2011 and others including Turi (acquired by Apple) and Algorithmia (acquired by DataRobot) shortly after. Fast forward to today, and and we’re proud investors in many next-gen intelligent apps like Seekout, Highspot, Amperity and Tesorio, to name a few, and those companies that enable intelligent applications like OctoML and WhyLabs. We also work closely with leading institutions and researchers at the University of Washington and Allen Institute for Artificial Intelligence, to understand emerging capabilities and to build innovative startups around these technologies. We also have close relationships with leaders and their teams at Amazon and Microsoft, two of the most innovative companies in the world both building and deploying intelligent applications and providing services for others to create them.
Intelligent Apps 40
This fall, Madrona Venture Group and Goldman Sachs collaborated on a new initiative to research, identify and select the top intelligent application companies in the private sector. The IA40 initiative enlisted judges across 40 of the very best venture capital firms a to participate in this research-driven ranking. Judges nominated over 250 companies that they believe will transform and define the next generation of software with application intelligence, and voted for the top 40 most promising.
Judges were asked to nominate companies and later vote across four different categories below.
- Early stage – up to ~$30 M raised
- Mid-stage – ~$30 to ~$200 M raised
- Growth-stage – ~$200 M or more raised
- Intelligent Application Enablers (stage agnostic)
- Intelligent App Enablers are the most clear and well-known group of companies. Several like Databricks, DataRobot and Fivetran have emerged over the last 5+ years as market leaders, and many of the earlier stage companies like OctoMl and dbt are already gaining significant market traction.
- Machine to Human interactions are starting to become more ubiquitous – as the categories of sales (Gong), Marketing (Amperity), Customer Success (Cresta, Moveworks) grow.
- Machine to Machine interactions continue to rise in adoption as cybersecurity platforms (Abnormal), physical security platforms (Anduril), and horizontal enterprise process automation (Celonis) reach the growth stage
- Cloud Providers: While the whole journey of intelligent applications (including model design and training) is hybrid and multi-cloud, most intelligent apps today leverage and operate on one or more of the major cloud platforms (AWS, Azure and Google Clouds)
- Investors: The investor group that has backed these Intelligent Apps largely represents some of the most venerable VC brands and current trends (for example, Tiger Global backed the most companies voted on to the list with A16Z and Sequoia tied for the second most companies)
Expectations for 2022
- Several IA Enablers will go public – as these companies see broader adoption in the market.
- There will be increased M&A and consolidation in the intelligent apps space – as both traditional software companies buy Intelligent Application companies (building on the Q3 2021 record referenced above) and consolidation amongst emerging market leaders in key sub-sectors.
- Assuming macro-economic conditions hold, 2022 will set another record year in VC funding for intelligent applications across all stages of private financings – we are still in the early innings!
- Data rights and regulations will increasingly impact strategies for IA companies – which can be both headwinds and tailwinds for individual companies as they attempt to responsibly access, use and leverage the fuel that powers intelligent applications.
- Flywheels will start to emerge in various sub-sectors of intelligent applications. The flywheel of leveraging diverse and robust data to create contextually relevant machine/deep learning models that are then deployed to help solve real-world problems and then the learnings from those inferences (which is more data) are incorporated in to further improving the intelligent applications will be a source of sustainable competitive advantage for successful companies. And, it will encourage usage and engagement go-to-market and pricing models that facilitate customer adoption.
Working Definition of Intelligent Applications:
Intelligent applications derive insights from data, and leverage machine intelligence to help solve a business problem better than ever before.
These insights come from first accessing data that lives in various data silos (often the systems of record underlying software applications) and aggregating it into a data lake or modern data warehouse. This data, once appropriately cleaned and prepared, can be used to train and test models that act as a representation of the questions or decisions a business user is looking to answer and make decisions upon.
Those models are trained using technical processes that we generally refer to as machine and deep learning. Once they have been trained and optimized to address the contextual issue they are intended to, they are eligible for deployment in production – which often means delivering results in real-time. Deployed correctly, these packaged end state applications generate contextually relevant predictions, recommendations, inferences or insights in a modern API and user experience. Finally, intelligent applications are enabled to continuously monitor and improve the underlying models and predictions they surface, based on the data they continuously input.
We believe that every successful and enduring new application built today and, in the future, will be an intelligent application.