Quick on the Uptake: Digital transformation for industrial companies
The future for Uptake is promising, but there are lots of opportunities – or potential pitfalls. (706)
Machine learning is being heralded as one way to help a growing company streamline data, improve processes, and eliminate inefficiencies in nearly every aspect of the organization, but developing in house solutions seems to be a challenge for smaller companies. With demand growing for these adaptive solutions, so too is the supply of companies eager to assist. Already identified as a leader in the field, Uptake Technologies Inc. (Uptake) is positioning itself as a master platform for machine learning that can transform a company across various industries. 鈥淏y harnessing the power of information to help companies make better predictions, Uptake turns every industrial company into a digital聽organization.鈥� Keeping this edge is driving Uptake.
A common concern for employing machine learning is predictive analysis versus causality. Companies within the machine learning space understand this concern and are working to take large amounts of digital information, turn it into useful data, and then develop algorithms necessary to find the otherwise indiscernible way forward. So far, investors show confidence in Uptake鈥檚 ability to deliver on predictive analysis. In November 2017, Uptake closed a Series D round with $117 million earmarked for growth opportunities. They also moved into the defense sector with their winning bid for the Defense Innovation Unit, Experimental (DIUX) in June 2018. Currently the Internet of Things (IoT) industry, which encompasses machine learning, is expected to be worth $70 billion in 2020. The military industry alone is valued to grow from $6.26 billion in 2017 to $18.82 billion in 2025.
Uptake is diversified across ten industries, so one can imagine they have had the opportunity to create rather large databases and test multiple variations of their algorithms. However, it is important to recognize that these are industrial applications. Uptake is not venturing in to the realm of healthcare or retail operations at this time. In the short term, I think it is prudent to remain true to the kind of company Uptake is: a leader in industrial machine learning. They have experience in this relatively new space and are a leading firm. This affords them certain benefits of being the firm for your machine learning needs (as long as it is an industrial problem). Securing this benefit for the long term is another issue.
Past performance is not a guarantee for success moving forward. While it would not be advisable to make a move into retail or healthcare at this time, I think it would be a misstep to not entertain this as a possibility. Securing dominance in Uptake鈥檚 current field preserves the funds necessary for innovation and development of forward looking capabilities. They must maintain their edge as the leader, which is no easy feat given the pressure from Google and Amazon, while developing applications to industries they have not yet entered. Moving to government contracts gives some stability to this fast-paced market.
The future for Uptake is promising, but there are lots of opportunities – or potential pitfalls. What kind of competition do they really face within the industrial applications? Will the next step be moving into retail? Will Uptake remain a business to business (B2B) firm? Can they acquire a company already operating in this space? There is always risk in growth, but in this world, stagnation means catastrophic failure. Whatever Uptake has up their sleeves, the machine learning field is waiting. (706)
Uptake Technologies Inc., 鈥淎bout Uptake,鈥� https://www.uptake.com/about, accessed November 2018.
M. Yeomans. What every manager should know about machine learning. 性视界 Business Review Digital Articles (July 7, 2015).
E. Brynjolfsson and A. McAfee. What鈥檚 driving the machine learning explosion? 性视界 Business Review Digital Articles (July 18, 2017).
Ingrid Lunden, 鈥淯ptake nabs $117M at $2.3B valuation for industrial predictive analytics,鈥� Tech Crunch, November 30, 2017, 聽[https://techcrunch.com/2017/11/30/uptake-nabs-117m-at-2-3b-valuation-for-industrial-predictive-analytics/], accessed November 2018.
Sydney J. Freedberg, Jr., 鈥淎I Logistics Let Combat Units Move Faster: Uptake鈥檚 DIUX Contract,鈥� Breaking Defense, June 27, 2018, [https://breakingdefense.com/2018/06/ai-logistics-can-speed-up-army-tanks-uptakes-diux-contract/], accessed November 2018.
Nicholas Shields and Peter Newman, 鈥淎mazon introduces host of new IoT services 鈥� Uptake eyes industrial predictive analytics 鈥� Google Home adds multi-command feature,鈥� Business Insider Intelligence, December 5, 2017. [https://intelligence.businessinsider.com/post/uptake-eyes-industrial-predictive-analytics-2017-12], accessed November 2018.
鈥淎rtificial Intelligence (AI) in Military: Global Industry Analysis & Forecast 2018-2025,鈥� Business Wire, June 6, 2018, [https://www.businesswire.com/news/home/20180606006201/en/Artificial-Intelligence-AI-Military-Global-Industry-Analysis], accessed November 2018.
Uptake Technologies Inc., 鈥淚ndustry Solutions,鈥� https://www.uptake.com/industry-solutions, accessed November 2018.
Sample Image from Uptake Technologies Inc., 鈥淧latform,鈥� https://www.uptake.com/platform, accessed November 2018.
To me, the Defense Innovation Unit, Experimental (DIUX) victory in June 2018 is the single biggest indicator of Uptake’s potential. As a Company ~12 months removed from their Series D round, Uptake should be focused on securing long-term, cash-flow positive contracts — the Department of Defense is the ideal customer from this perspective. Given the knowledge Uptake will gain working with the DoD, I would advise they seek to be the dominant force in the Industrials (i.e., Defense, Aerospace, Chemicals, Metals & Mining, Packaging, etc.) sector. Retail and Healthcare both seem to be risky, and concentrated, sectors to penetrate next. With regard to Retail, Amazon et al already have a tremendous amount of data and engineering skill that make it look like the clear winner in Machine Learning Retail. Moreover, secular headwinds in Retail make it a very risky proposition. With regard to Healthcare, I worry that the complexity and ever-shifting nature of Healthcare regulation in the United States and abroad make this end-market a difficult to win space – healthcare customers tend to be of poorer credit quality. You mentioned IoT and I believe that working with Telecom companies to be the Machine Learning player in the next wave of 5G with telecom companies is a better risk-adjusted bet if the Company must find an adjacency.
Agreeing with CPM, Uptake is in its unique industrial space and still has a lot more to explore within the field. Moving into retail or healthcare may require more domain expertise to out run the current players in the field. Uptake recently underwent a mass layoff and reorganization which may not be a good time to introduce a radical change in the company products. Instead, they can try to leverage the DoD contract and penetrate further into homeland security and forensic science.