ServiceNow Knowledge 2024: New Generative AI Features Unveiled

In May 2024, a 20,000 strong ServiceNow community descended on Las Vegas for 3 days of breakout sessions, training, thought leadership, networking, and partying at Knowledge 24. Knowledge is ServiceNow’s flagship annual conference for customers, partners, and advocates. Although the event is typically centred around customer stories, there were several exciting new announcements that we’ll break down in this blog.

You can find out more about the event and watch the key note sessions (no need to login) on the Knowledge 2024 site. A more formal news roundup can be found on the ServiceNow Announcements page and the ServiceNow Blog.

Here are my top 4 highlights from Knowledge 24:

1. Showcasing Generative AI!

In the opening key note on day 1, Put AI to Work for People, Bill McDermott spoke about the impact of Artificial Intelligence (AI) on the workforce. We saw transformation play out in an on-stage bakery with Omar El Gohary, ServiceNow VP for IoT Solutions (14:20). Existing information from point-of-sale systems, IoT-enabled equipment and sensors, enterprise software, and other data sources, was integrated into a single platform using Integration Hub. Thanks to the platforms baked-in GenAI and automation, the company was able to increase growth, predict demand, and reduce wastage.

The session was packed with more real world customer stories from Aston Martin, Warner Brothers, HSBC, and others. BT Business for example (27:45) have seen a significant improvement in their Net Promoter Score (NPS), their incident Mean Time To Resolve (MTTR) has halved, and they realised a 90% improvement in incident predictability. Furthermore, as a service provider, many of these benefits are passed on to their customers, with one requiring 80% less manual work for processes, and another achieving 75% automation of network and security changes.

Throughout the key notes GenAI was put to work, there were multiple live demos in Unleashing the power of generative AI with the ServiceNow platform with Joe Davis, ServiceNow SVP for Platform and AI Engineering. First up a new input type has been added to Now Assist, you can take a picture! Now Assist Image to Workflow identifies the image, in this example a process flow for device upgrades drawn out on a whiteboard, and converts that to workflow constructs within ServiceNow in seconds (23:10). Further demonstrations of customer and agent experiences with Now Assist followed, including automatically generated incident resolution notes and Knowledge Article Generation to supercharge agent productivity (26:05).

Bring Your Own (BYO) GenAI Model with Now Assist is now supported to compliment the existing multimodal options. The BYO GenAI model allows for flexibility and customisation of AI-driven transformation for larger organisations. Post Call Summarisation uses GenAI to summarise transcripts of voice calls to reduce call handling time and manual effort from customer support agents. This feature is available initially in Now Assist for Customer Service Management (CSM).

Now Assist has added External Connector Support, meaning it can search everything across the enterprise to provide answers using both ‘internal’ ServiceNow data and ‘external’ data sources. In the demo we saw a Sharepoint connector activated, which then featured in an employee portal search for health benefits (50:00). The result doesn’t just link to articles, which might take an employee 5 or 10 minutes to find the answer, it gives the answer directly in seconds.

Now Assist for Strategic Portfolio Management (SPM) was announced to connect customer feedback with strategic planning. Using GenAI, SPM gathers and analyses feedback and requests in a single place, driving business impact and decision-making. A new conversational experience for demand creation also allows business users to leverage natural language to create and modify demands, rather than filling in lengthy intake forms.

2. Build me an App!

New Creator Studio, available with App Engine, introduces a no-code development studio and automation capabilities enabling anyone to build applications on the Now Platform. Using ServiceNow’s GenAI Now Assist, people can Generate Applications, Playbooks, and Catalogs using natural language from an intuitive user interface.

What’s critical here is that non-technical employees can build applications, improving productivity and empowering innovation from the front line. During the CreatorCon key note Jithin Bhasker, ServiceNow VP & GM for App Engine, took to the stage to demonstrate the power of application generation, building out a coffee ordering app in minutes entirely from natural language prompts (16:11).

For developers, the same flow can be opened in ServiceNow Studio represented as code (coming in the Xanadu release). And whilst the platform provides low and no-code capabilities, the Integrated Development Environment (IDE) allows developers to build in code for advanced automation and integration. The IDE visualises code changes side-by-side, pro-actively catches errors at build time instead of run time, and introduces Developer Sandboxes. Developer sandboxes mean that multiple developers in the same team can work on the same instance, same flow, same app, and same code base, but within their own branches.

3. AI Friends!

At Knowledge 24 ServiceNow announced collaborations with Microsoft, NVIDIA, IBM, Equinix, Fujitsu, Infosys, and Genesys. Let’s look at the first two in more detail.

CJ Desai, ServiceNow President and COO, hosted the day 2 key note Unleashing the power of generative AI with the ServiceNow platform, and was joined by Srini Raghavan Microsoft VP of Product for Teams, to showcase the joint engineering work the ServiceNow and Microsoft teams have been doing to integrate Now Assist with Copilot (9:40). The partnership makes it easier for employees to execute common productivity tasks.

During a live demo, Srini walked through a scenario of an employee returning from leave, catching up with summarisations of what they missed, and ordering a replacement laptop after the AI pro-actively identified an underlying issue. The Copilot Teams chat integrates and seamlessly hands off to Now Assist, enabling the user to self-serve and prevent a loss of productivity through device outage.

Jensen Huang wants Service, Now! Jensen, NVIDIA Founder and CEO, joined CJ on stage to talk about ServiceNow’s AI journey (53:45). GenAI has been built into the Now Platform since 2023, but the foresight started much earlier than that. Jensen recognised ServiceNow as the first enterprise software company to start working on a full stack generative AI platform. The two companies had the same vision early on, and worked together on building, training, and optimising AI models.

We got a look at the future of Now Assist with a demo of AI avatars of the future, leveraging with ServiceNow language models and NVIDIA’s Avatar Cloud Engine (ACE) digital avatars. The example brought digital characters to life using voice to text, and text back to voice, to hold an end to end conversation with a customer and resolve an Internet service issue.

Here are 3 great soundbites from Jensen:

• Over the last 10 years AI has scaled a million times. If something is moving a million times faster every 10 years, what should you do? The first thing you should do, instead of looking at the train from the side, is get on the train! When you’re on the train, it’s not moving that fast!
• The reason for the pace of AI adoption is that we’ve transitioned from instruction driven computing, which is coding, to intention driven computing, which is intelligence. While fewer people can code, intelligence such as Now Assist can be used by anyone, as people don’t need to describe a problem exactly to find a resolution.
• Most AI now is one-shot, you prompt it with something, and it instantly generates a recommendation. In the future, AI will be multi-shot, using a multi-step reasoning based system. This will lead to generalist and specialist AI’s using other tools and AI’s working with other AI’s, in much the same way as employees work with other employees.

4. Improving Employee Experience!

First up is a future roadmap item; ServiceNow are redefining the employee experience with My Assist, a personalised assistant for work built on Knowledge Graph (46:20). Knowledge Graph is a new technology, it captures your processes, integrations, requests, and assets and leverages that information to form context-based intelligent workflows.

Using Knowledge Graph, My Assist builds a personal and cohesive omnichannel engagement layer for each individual. The experience layer is a people centric, meta-data driven framework, that is tailored to the individual and combines services (such as software and hardware), integrations (such as payroll, expenses, holiday systems), and staff (such as management, team-mates, collaborators, approvers).

During the day 2 key note, we saw a comprehensive onboarding demo for a new employee, using My Assist on the mobile app through to transitioning onto a laptop device during the first day in the office (48:10). My Assist provided a personal assistant for all the new starters administrative and role-based needs, presumably reducing their ramp up time to productivity!

In other updates, ServiceNow Talent Development was expanded to identify skills gaps, training needs, proficiencies, mentoring opportunities, and more. Talent Development integrates with Manager Hub and Now Assist for HR Service Delivery to enable a holistic view into the training and development of individuals and teams.

Workplace Service Delivery also benefits from an injection of AI, with a chat-based concierge experience for common requests such as booking meeting rooms, and recommendations on when teammates are in the office to maximise collaboration. Smart building integrations with leading sensor companies will also deliver real-time, actionable insights on building occupancy and experience.

One way to improve employee experience and productivity is to reduce day-to-day IT outages. Digital End-User Experience (DEX) ensures real-time visibility across devices and applications, logs user trends, and proactively identifies problems. By linking employees desktop experience with the Now Platform IT can minimise outages, speed up resolution, and improve technology experiences. In some cases, employees create inefficient workarounds for issues rather than contacting IT. With DEX, IT already know about device issues, often before the symptoms have disclosed themselves to the end user. DEX will start with application and device health before an AI-powered desktop assistant becomes available later this year.

In another notable customer story Jacqui Canney, ServiceNow CPO, spoke with Warner Brothers about creating an empowering employee experience for story tellers, following the 2022 merger with Discovery (22:35). Warner Bros implemented a chatbot called Ask Sheldon to quickly provide answers across IT, HR, and finance. This gave employees a consistent, streamlined experience, reduced agent time, and enabled cross-departmental collaboration to blossom.

And Finally!

Just because I can’t miss these out, Pat Casey, ServiceNow CTO, spoke about the platform initiatives focusing on scale, performance and security (4:30). Here are some of the cool announcements:

• RaptorDB is an entirely new back-end database for both transactional and analytical workloads, that will drastically improve processing speeds, database queries, and CMDB list load times. Before RaptorDB a typical ServiceNow instance could achieve 35,121 transactions per minute. After RaptorDB 93,257 transactions per minute were logged, resulting in a 3x workload.
• Faster upgrades starting in Xanadu, ranging from 26% faster for smaller organisations up to 75% faster for larger organisations. Turn on new features in a fraction of the time.
• Faster workspaces mostly in Xanadu; 41% faster load time for IT Operations Management (ITOM), 53% faster for IT Asset Management (ITAM), 33% faster for Customer Service Management (CSM), and an average of 25% faster load times across all workspaces.
• Better security with new Query Access Control Lists (ACLs), Deny-Unless ACLs to define exception cases, and highly Granular ACLs for existing table-based ACLs to define tables and rows with new ‘applies-to’ conditions.

In conclusion, Knowledge 24 delivered excitement and customer wins. Naturally the event was GenAI heavy, but the application of AI will start with small and direct improvements, like reducing manual work for service agents. Organisations can make quick and easy wins to start permanently putting out some of those fires before continuing on the journey of transformation. If you found this interesting and have more time, this post-Knowledge Forbes article is worth a read: ServiceNow’s Bold Gambit To Become The Platform Of Platforms. Start thinking about attending Knowledge 2025, it promises to be even bigger and better!