Pioneering progress: ServiceNow's journey with generative AI

With GenAI quickly transforming the way organisations operate, we dig into ServiceNow’s history with AI and the promise GenAI holds for the future of the Now Platform.

Artificial intelligence (AI) and machine learning (ML) have been on the scene for many years, but the arrival of GenAI has changed the game in how consumable AI can be with its ability to produce novel and ‘human-like’ outputs.

Organisations of all sizes are now excitedly exploring how GenAI might enhance and transform their business models and services – and ServiceNow is helping lead the charge.

This blog will explore ServiceNow’s journey with AI, the new GenAI features you can adopt right now, and make some predictions on how GenAI will shape the Now Platform in the near future.

It’ll be interesting to return to this in a couple of years and see how wrong (or maybe right) I was!

ServiceNow and GenAI

ServiceNow is a hugely innovative organisation, as shown by its heavy investment in R&D and global innovation centres. So, it's no surprise that AI and machine learning within ServiceNow’s platform aren’t new.

For many years, ServiceNow has actively integrated AI and machine learning capabilities into its platform to help organisations automate processes, improve efficiency and provide better experiences for users.

Existing AI capabilities in the Now Platform include Virtual Agent, Predictive Intelligence, and AI Search to name a few.

What’s exciting now is the inclusion of GenAI and making AI capability more accessible and consumable to all within the platform.

And, to take it even further, this isn’t some quick fix integration, we’re talking about a full partnership and domain-specific AI for each customer!

AI in ServiceNow

ServiceNow has pioneered the use of AI within its service management and enterprise workflows since 2017, acquiring several AI solutions that have now become staple elements of the Now Platform.

Where many organisations simply bolt products together, ServiceNow has impressively re-engineered its acquired AI capabilities to work natively in the platform.

Below is a timeline of ServiceNow’s AI acquisitions to date and the capabilities implemented as a result:

2017

• DxContinuum: Machine learning for process automation
• Qlue: Virtual Agent powered by AI

2018

• Parlo: Natural Language Understanding
• FriendlyData: Natural Language Query (Performance Analytics)

2019

• Attivo: AI-based search

2020

• Loom Systems: AIOps
• PassageAI: Multi-Language AI
• Element: Enterprise AI (NOW Intelligence)

2022

• Hitch Works: AI Talent & Skills Mapping (Employee Workflows)

With its recent announcement of partnerships with NVIDIA and Hugging Face, ServiceNow is quickly increasing the scale of its AI innovation and investment.

Accessing GenAI in ServiceNow

At Knowledge23, ServiceNow formally announced its plans for introducing generative AI into the platform.

These plans include two ways for adopting GenAI within your ServiceNow platform:

1. Using ServiceNow’s ‘Generative AI Controller’ to integrate with available GenAI providers – i.e., OpenAI and Azure.
2. A domain-specific Large Language Model (LLM) created in partnership with NVIDIA to deliver ‘Now Assist’.

ServiceNow’s partnership with NVIDIA will use its state-of-the-art language model system. This includes pre-trained models of large, medium and small sizes, as well as training between family models – large training medium, medium training small, small training tiny, for example.

It also includes guard railing systems, supervised fine-tuning systems, reinforcement learning and human feedback systems, and vector databases that will connect it with your proprietary data and information.

The collaboration between all these elements is what allows it to create specialised models for you.

The domain-specific approach

Whilst the Controller integration will likely be a faster route to adoption, security posture and risk appetite should also come into consideration.

In the domain-specific approach, NVIDIA will utilise their NeMo development framework to allow customers to create a domain-specific LLM for more accuracy, optimised data and data privacy. Not even ServiceNow will have access to your LLM/data.

The domain-specific approach will likely produce three main advantages:

1. Relevance: You can train your domain-specific data to create outcomes that use context and words relevant to your organisation.
2. Trust: You can control guardrails for inputs, output toxicity and bias threshold, all relevant to your data – avoiding undesired responses or ‘hallucinations’.
3. Control: Your users can train your model with their interactions – creating a model that’s built solely for your organisation.

Plenty of benefits – but it’s not a shortcut to success

ServiceNow's Generative AI Controller and Now Assist solutions will give users more specific answers to questions as opposed to producing a knowledge article or link, and do so in a more personalised and empathetic manner.

And the benefits don’t stop at users. Agents will be able to have a contextual summary of record histories instantly – no more scrolling through work notes and activity panels.

They’ll be able to conduct more robust root cause analysis and create/update accurate knowledge. And developers can enhance the platform using text-to-code, text-to-flow and text-to-app interactions.

Your ability to grow and scale the platform is more efficient than ever before, meaning more innovation and greater business results from your investment in ServiceNow.

But here’s the catch...

It's not an automatic route to innovation or productivity improvements. You can’t cut your way to innovation (as said by Jony Ive at Apple). You will need a strategy for adopting AI within your organisation, and analysing your current vs. future employee skills, processes and data will be essential.

The most successful companies will use GenAI to reimagine roles and augment and empower their employees to outperform everyone else.

ServiceNow’s GenAI roadmap

Whilst the ServiceNow roadmap is something of a crystal ball until it arrives, we can take recent news as a guide to its likely direction.

In May 2023, ServiceNow released the Generative AI Controller v1 and Now Assist for Virtual Agent, but ServiceNow’s wider GenAI capabilities will become generally available in its Vancouver release in September.

This will allow you to adopt Now Assist for case summarisation and text-to-code and is compatible with all platform workflows to deliver conversational experiences across all use cases.

There’s so much potential, but here are a few predictions for how ServiceNow might evolve GenAI in the coming months (based on the time of writing and factoring in current industry trends):

• Integration of voice commands with GenAI to create reports, summaries, applications and forms.
• Advances in long-context tasks such as multi-document understanding and summarisation (document intelligence).
• Greater use of GenAI in more strategic perspectives, i.e. ESG and Strategic Portfolio Management.
• More assist capability for app creation, flow designer suggestions, code creation and deployment.
• New algorithms to improve task balancing and performance.
• New approaches for large-scale data generation for training AI.
• Translation improvements using language models.
• Amortised optimisation to predict the outcome of an optimisation process.

Conclusion

All in all, the future is bright for ServiceNow users. As we said, it’s not an automatic win and will require some careful thought – but there’s no denying that ServiceNow’s journey with GenAI holds exciting potential.

We’re following new developments and features as they happen, so be sure to subscribe to our newsletter to make sure you don’t miss those (particularly as the Vancouver release fast approaches!)

Alternatively, if you’d like to find out how to get started with Generative AI Controller or Now Assist, just reach out and we’ll help you explore how your organisation stands to benefit from ServiceNow’s latest AI innovations.

Key takeaways

• GenAI holds a lot of potential for the future of the Now Platform.
• ServiceNow have already implemented GenAI in the form of Generative AI Controller and Now Assist.
• ServiceNow has a long history of investing in AI to improve its native platform capabilities and is quickly looking to innovate at scale with new GenAI partnerships, such as Nvidia and Hugging Face.
• GenAI will be a powerful tool in improving life for users, agents and developers – but it’s not a shortcut to success and requires careful consideration if you’re to get the most out of it.

As an elite ServiceNow partner, FlyForm is renowned for its specialised skill in ServiceNow implementations, customising and refining ServiceNow solutions to meet the specific needs of businesses, ensuring a seamless digital transformation. Our expertise in ServiceNow App Creation amplifies IT capabilities, empowering businesses to enhance their operations through tailored apps and automation, leading to greater efficiency and streamlined workflows. We have a proven track record in elevating IT Service Management (ITSM) solutions through strategic ServiceNow implementations, assisting businesses in harnessing the full potential of ServiceNow ITSM by automating processes and integrating disparate tools into a unified system.