Date: Jul 08 2020
We have come a long way from the days when Artificial Intelligence (AI) was primarily a research project in academia or in the corporate world, and the size of the market opportunity was unclear. Today, the transformative power of AI is well recognized and the dollar value generated by AI is estimated to be at least a $3 trillion dollars annually (through sales, cost savings, automation, etc.) according to McKinsey.
While academia places its focus on general AI, it has been narrow AI, which is designed to handle specific tasks, that has gained adoption and risen as a highly investable space with strong returns for customers and investors alike. On the technology side, a coalescing of trends including the pervasiveness of sensors, cheap storage and compute, and algorithmic breakthroughs have enabled the adoption of narrow AI.
Another key factor for narrow or applied AI to materialize is the availability of (good) data at scale. With the massive increase in data generation and enterprise companies’ demand for using it, the need for a new smart data infrastructure that can perform at a greater scale has arisen. Disruptive startups like ChaosSearch, a data lake engine for log analytics at scale, are building the intelligent backbone that can power a machine learning-enabled world. The increasing demand for good data has also proven to be a key challenge that savvy AI startups have used as an entry point into the enterprise. Even today, most companies have their data siloed and spread across disparate databases, in different formats and not at enough scale to train data models that can leverage AI. Savvy human augmentation startups such as Zylotech, a pioneering self-learning Business to Business (B2B) Customer Data Platform, has seen this problem as an opportunity. Zylotech started by “cleaning” and enriching companies’ data, getting its early customers to subsidize their R&D efforts and the creation of its proprietary dataset. Today, their self-learning customer data and analytics platform provides continuous information updates to improve data quality, personalization and actionable insights for marketing, sales, and data operations teams at Cisco, Dell, Palo Alto Networks, and Google.
Initially, AI startups focused on enterprise departments that were already big data adopters, such as marketing, IT, and cybersecurity. Both buyers and users in these departments had exposure to the advantages of being data-driven and had better quality data, enabling AI startups to deploy their products faster and demonstrate the potential of AI-first solutions. However, with AI delivering value across use cases, its adoption has expanded beyond these early adopters. Traditional industries such as manufacturing have realized the potential of their data, giving rise to the emergence of intelligent vertical platforms. Startups such as Verusen, an intelligent supply-chain optimization platform, are purpose-built for specific industries and leverage their specific datasets to provide intelligent recommendations that are tailored to the needs of that industry and have quick and measurable business value, which is helping to drive adoption. To exemplify the impact of Verusen’s AI platform, in less than 100 days, the startup generated more than $20 million in verified savings for a Fortune 500 pulp and paper manufacturer. Leveraging AI/machine learning, Verusen’s platform provided a simple, low-touch, self-cleansing approach to structuring materials data, allowing the company to accelerate its global sourcing and procurement strategies.
Frontier Tech and AI Adoption Acceleration
The impact of COVID-19 has accelerated digital transformation and AI adoption. As the pandemic persists, we are seeing a divergence between technology and non-technology sectors. This is clearly visible in the relative divergence in performance between the NASDAQ, which has rebounded to all-time highs, and the S&P 500, which has yet to rebound to this year’s February levels. The pandemic has changed societal norms, normalized digital interactions, and accelerated pre-existing trends of cost-cutting and digital transformation. We have seen 10 years of digital adoption across industries accelerate overnight, providing added momentum to the AI wave. As companies become more digital, AI becomes indispensable in driving cost efficiency and superior user experiences at scale. It is no surprise that while the majority of sectors in the economy are experiencing significant reductions in demand, cloud computing adoption is expected to see 19% growth in 2020, digital transformation is expected to experience 10% growth this year, and 44% of CFOs surveyed by PwC plan to accelerate their spending in automation. These trends are mirrored in our own portfolio. As non-tech companies across various sectors are adjusting and bringing down costs through automation and other technologies, our founders and their startups have managed to retain all their customers with the vast majority growing at, or above plan. Put differently, what the current environment has demonstrated is that in good and (extraordinarily) difficult times, the digitalization of the economy and the ubiquity of AI are here to stay.
Domain Experience in Early Stage AI Investing is Key for Superior Venture Returns
From an investment standpoint, as the economy experiences this massive technological shift, not investing in technology means being left behind. In fact, in our view, it is AI and advanced or frontier tech investing that will drive outsized results in venture capital because of the transformational and measurable power of AI-enabled products and platforms.
When investing in AI, it is important to understand the differences in how these startups are built. This is where the domain experience and expertise of early-stage AI investors play a critically important role. Firms like our own understand that AI startups may require more time and upfront costs to get to market because they need to train their data-powered algorithms. Furthermore, we understand that as the required data is usually hosted in customers’ existing enterprise applications, it is essential for AI startups to assess how they fit within target users’ software toolkits and workflows, and how to build the required integrations to maximize value and minimize behavior change. As data models are at the core of AI products, product development teams need to incorporate data scientists, who have a fundamentally different culture from software engineers, which increases management complexity.
At Glasswing, where we have domain expertise in early-stage AI investments, the market opportunity for portfolio companies is large and the companies are defensible. Once the core dataset and data acquisition strategy are established, these AI startups’ solutions can tackle use cases that are impractical for typical software startups. Their output is more generalizable and keeps improving with additional data. This creates the data network effect, which continuously increases in value to customers.
AI Wave – From Foundations to Broad Adoption
AI is here to stay, and the opportunity to invest in AI is ripe in these early days. We are moving from creating the foundations of AI to democratizing access to its power across applications. The current environment, along with recent trends makes AI now more essential than ever, particularly in the enterprise. I am excited to have a front-row seat by backing pioneers in this space, who are automating the mundane and unleashing human potential across the economy.
Rudina Seseri is the founder and managing partner of Glasswing Ventures, an early-stage venture capital firm dedicated to investing in enterprise, platform, and security startups that harness the power of AI and frontier technologies to transform markets and revolutionize industries. Hear Rudina speak on Harnessing the Power of Artificial Intelligence at FOX's Virtual Global Investment Forum, July 14-17.
Email: email@example.com | Website: glasswing.vc | Twitter: @rudina11