Joint post with my colleague Holger Mueller (@holgermu)

Element AI Takes Workflow to the Autonomous Level for ServiceNow

On November 30th, ServiceNow announced an estimated $500 (CAD) million acquisition of Montreal, Canada based startup, Element AI.  Co-founded by Dr. Yoshua Bengio, a 2018 winner of the Turing Award, Nicolas Chapados. and Jean-François Gagné in 2016, the startup's mission was to bring AI to non-tech companies in order to compete with tech based companies (see Figure 1).  Key products include an AI-assisted insurance underwriting workflow software known as Underwriting Partner and a data set management platform for manufacturers known as Knowledge Scout.

Figure 1. Jean-François Gagné, co-founder and CEO of Element.AI

Source: Element.AI

In a blog post, Element AI"s co-founder and CEO noted “Element AI will help ServiceNow deliver workflows that learn more efficiently from smaller datasets, improve the quality of existing AI capabilities like content and language understanding, and expand new capabilities like image recognition and cognitive search."  Gagné also pointed out this level of integration would help users “summarize information, make predictions and recommendations, and automate repetitive tasks,”

As ServiceNow's fourth AI acquisition in 2020, CEO Bill McDermott and Chief AI Officer, Vijay K Narayanan, has doubled down on the investment and future of the platform with the build out of Now Intelligence capabilities along with the acquisitions of Loom Systems, Passage AI, and Sweagle.

Acquisition of Element AI Elevates ServiceNow Into The AI Driven Orchestration and Automation Market

AI changes enterprise automation holistically with its ability to automate manual labor through software. For ServiceNow, that is on the quest to become the universal workflow platform for the enterprise, it is the game changer that will attract CxOs to its platform. Any disruptor needs to show automation benefits, and delivering automation from AI can the difference maker for enterprises to adopt and migrate to new workflow platforms. And even for existing customers, in the ServiceNow areas of IT, customer and employee workflows, AI is a key change agent for a better automation future. And enterprises that automate more efficiently and become more agile, practice Enterprise Acceleration, the key fuel to enterprise survival in disruptive and tumultuous times. 

1. Element AI addresses several product gaps and creates key opportunities - On the functional side ElementAI brings strong productized document processing to ServiceNow. It has good document processing capabilities that are part of its Document Intelligence product and on top of that its Knowledge Scout product creates knowledge models and workflows. Access Governor applies the knowledge from documents and recommends role based access for the enterprise. All three offerings provide immediate value for ServiceNow’s customers.

   Element AI has done a good job also monetizing its APIs to enterprises, an offering that is less likely to remain around now with ServiceNow, as we don’t expect the vendor will show interest in regards of AI related advisory services (beyond the immediate product scope of ServiceNow). Of substantial value will also be ElementAI’s OS platform - if ServiceNow manages to empower a broader range of users to build AI based next gen Apps. If Orkestrator is of value remains to be seen - given ServiceNows partner strategy, it is more likely the vendor will most likely use its cloud partner’s platform and orchestration platforms. 

2. Element AI is an acqui-hire - Talent plays a key role in the battle for AI, and ElmentAI brings key talent to ServiceNow - effectively soaking up most of Canada’s AI royalty from both a research and commercial perspective. Given Canada’s substantial role in deep learning is a good move as well, but ServiceNow needs to find ways to keep the talent on board. ServiceNow may face some challenges retaining the team as AI talent is in high demand.
3. Documents lead the way to workflow - ServiceNow has understood that documents are the artifacts of the digital enterprise - and are like the start and finish locations of a cycling stage race… with workflow the race between the start and end points. A lot of workflow can be derived from documents and that is where the Element AI IP comes in handy, with its document centric capabilities.

Customers Finally Gain A Future Roadmap Beyond Workflow

While many industry watchers would agree that the move to embrace workflow as the key messaging point has helped shift the perception that ServiceNow is both more than an ITSM offering and a key cloud platform. However, the larger differentiator in the next 18 months will come from both automation and AI.

1. ElementAI goes from solution to ingredient - Element AI customers have found a viable future for the vendor that was in need to find another funding round. But the concern will be what will happen with offerings and services that are outside of the likely interest of ServiceNow. Customers relying on these offerings and services need to get reassurance from Element AI asap - better even from ServiceNow. Understanding strategy, product roadmap future and possible end of life plans and deadlines will be key.
2. ServiceNow gets more heft in AI - It’s good to be a ServiceNow customer right now, as the vendor is expanding its automation capabilities and enterprises get more value from a proven / trusted platform. It also helps that IT in most cases runs it and there is no gatekeeper role here for LoB deployments - au contraire we see IT pushing new ServiceNow offerings. Element AI brings critical capabilities to ServiceNow, and now it will be all about how ServiceNow will shape the future of workflow. AI is certainly a key part of that and ServiceNow’s ability to execute is now significantly increased. But acquisitions are one thing, delivering on roadmaps is another - and that’s what ServiceNow customer need to keep an eye on. 

Platform Considerations Highlight The Challenges Ahead

ServiceNow follows a multi-cloud strategy with partnerships available or in the making wth the three big cloud providers. Element AI has catered more to the on premise work of data scientists, so there is no conflict here, as it makes its solutions portable, and likely to be easily integrated into ServiceNow’s portfolio. 

Like all enterprise application vendors, ServiceNow needs to make a few key decisions on platforms when it comes to AI.  Constellation sees three crucial areas that need to be address:

• Data gravity is real. AI needs massive amounts of data, but moving data is expensive.  Bringing the AI closer to the data is a key design point. With cloud data egress costs being punitive enterprise need to understand and plan where their data will reside. 

• Customers need an abstraction layer to power AI. The heavy lifting is too complex for most enterprises and there are not enough data scientists out there. The democratization of AI is key for enterprise application success with AI and vendors need to provide a common layer for non technical users to own their automation destiny. 
• Model migration strategy must be addressed. When models cannot be built where the data is, the need to be migrated and tested with the new data. Ideally this is avoided, but for most vendors still a reality.

Bottom Line: The Race To Business Graphs Is Here, Automation and AI are The Facilitators To Decision Velocity.

 The convergence of workflow, process mining, robotic process automation, integration services, microservices, and low-code/no-code platforms drive the future of software.  This next battle in enterprise software will be the creation of business graphs.  Like social graphs which use social networks to provide signal intelligence and digital feedback loops to accumulate massive amounts of data that is mined by AI, business graphs will accomplish the same for enterprises. 

In the case of the enterprise, the relationships among users, documents, business processes, and contextual data will power the signal intelligence and digital feedback loops.  As the majority of data is collected by digital feedback loops via automated and ambient collection, these systems can improve their precision decision capabilities.  Automation and AI are the tools that bring scale to creating decision velocity.