UNFAIR ADVANTAGE

We are witnessing the start of the ‘cognification’ of the enterprise. Over the next few decades, AI will be infused into every product, service, process and experience. It will become table stakes across each area of large organisations – from Customer Service and Sales on the frontline to Finance, Legal and HR behind the scenes.

But what does this mean for today’s enterprise in the emerging Age of AI?

We have seen AI rear its head before, only to retreat again with no viable use cases. Unlike previous ‘AI winters’, this time the emergence of AI is different – leaving the domain of academics, scientists and mathematicians and entering the realm of engineers and designers. 

There is no longer the conundrum of whether to buy or to build. As more and more stable foundational models emerge, enterprises can move quickly to create their own AI and solve their real-world problems. Whether that’s fixing an age-old, labour-intensive process. Taking on repetitive tasks to create more meaningful jobs for employees. Or continually delivering better, deeper, richer customer experiences. With every successful AI initiative, another process becomes far smarter than before, and the speed and scale at which this occurs determines the degree of cognification of the enterprise.

Yet as any student of innovation will know, during previous technology revolutions there have always been the early adopters and the laggards. Some will naturally want to wait until every kink has been ironed out. While others will focus on the competitive advantage that comes with speed of action. 

Plenty of enterprise leaders are already demanding, “Get me value from AI this year.” Right now, the challenge is learning how to avoid the hype cycle, build what is valuable and reap the rewards of cognification ahead of others.

It is undeniable how much the world has changed in the past few years. Work is a delicate balance of remote or in-the-office models. Data continues to grow at an exponential rate (running at hundreds of terabytes per day). And compute power has now got to the point where we can train models on the entire Internet. Meanwhile, knowledge remains a primary source of competitive advantage for any enterprise. Which is why cognification through AI can add value from all areas of the organisation. Not simply by improving manual workflows but by enhancing most forms of knowledge work too.

Despite the historical disappointments of AI the world has moved on. Gone are the brittle automation attempts, underwhelming results like those of Robotic Process Automation (RPA) or statistical models that require constant human intervention. The availability of foundational models like Claude, GPT-4 and others represents a significant shift by addressing the failures of the past, enabling lots of different people to build models and fuelling a growing excitement for the future. 

Fundamentally though, any AI model relies on information. And it is hard to imagine an enterprise process that doesn’t involve data – from handling customer complaints to performing a market analysis to running a logistics operation. The goal within enterprises is to surface the intelligence that will improve internal processes, lead to the best business decisions, enhance customer experiences or reveal innovative products and services. While AI can extract value in these areas by itself, the step-change lies in hybrid intelligence.

Hybrid intelligence is the outcome derived by humans and AI working together more seamlessly. With humans creating data for the AI and the AI delivering insights and automation for humans. As the potential for cognification within an organisation grows, there is a need for people and machines to work in tandem to lift, maximise and speed up the intelligence available where it is most needed. Intelligence that can form the basis of those significant efficiency gains, leaps in productivity or grand innovations that will keep the enterprise ahead of its rivals in the years to come.

AI will sift, sort and suggest from a huge amount of data. Still, people will need to train the models, check for biases and take responsibility for the outputs. This interplay between human and machine intelligence is complex and nuanced. To find advantage through AI, you must be able to find rapid pathways to cognification and deliver hybrid intelligence within the context of your own enterprise.

If there is already plenty of excitement around the potential of AI within the enterprise then there is also a healthy level of caution. Cognification sounds good in theory but there can be many barriers to putting it into practice.

All of the hype surrounding AI suggests it’s easy to dive straight in. That may be the case for some smaller businesses but things are not so straightforward for the enterprise. There are lots of choices to be made before realising value. Some of which include:

• Who needs to acquire new skills
• Which datasets to rely on
• What AI models to use
• How to avoid adding risk
• Where to focus security efforts 
• When to add the appropriate guardrails

Making the right choices – and quickly – positions the enterprise to quickly find advantage through AI. However, it’s essential to build and explore while making these choices, not just talk about it.

There are countless instances of how hybrid intelligence could deliver value throughout an enterprise. To illustrate how it is actually delivering value, here are three examples that are already in production within large enterprises.

In most departments, there are multiple systems for similar processes as well as the inevitable interdependencies. HR is no exception. Navigating these systems to answer employee queries wastes time. 

By cognifying this process, employees now simply type in a natural language question – like “How will my tax deductions change if I move up a pay grade?” – and the AI knowledge agent generates an answer for them from lots of disparate systems. This deflects the multiple calls, emails and ad-hoc reports that take up so much of the HR advisor’s time. It also frees them to take on higher-value work like developing new workplace policies, ensuring compliance or recruiting, onboarding and resolving disputes.

Combining people’s expertise with the AI knowledge tool is delivering added value through hybrid intelligence. For example, by training the model to anticipate specific employee requests and then to become self-learning and self-correcting. But also by allowing HR advisors to focus more on the happiness of employees and living the values of the brand.

Enterprise contact centre reps (CSRs) spend most of their time on the phone. But they are also required to summarise these calls, categorise key talking points, identify any required follow-up actions and input that data into customer relationship management systems. 

Cognifying this workload through an AI call summarisation model frees CSRs to spend more time with their customers. The speed with which AI handles the call transcripts, categorisation and data ingestion is increasing throughput within the contact centre. Getting the AI to do the work with 100% accuracy also means the department can handle more – or more complex – calls while reducing wait times and the associated frustration among CSRs and customers. 

New agents often struggle to know how to answer questions or where to go for the right answers. They take time to triage and identify how best to serve the customer. 

Through cognification, they are presented with a combination of internal articles and previous agent responses by surfacing pre-existing transcriptions and tickets. The AI model summarises the information based on a prompt that reflects the customer query. Relevant information is then directed into the new agent’s workspace so they have appropriate answers in real-time. This means new employees can resolve issues faster. It also means that longer-serving employees are not called in every time a new agent needs assistance and can instead put their experience to work dealing with edge cases.

But simply building the initial AI model and expecting it to do all the work is not really viable over the long-term. What happens if 20% of what it returns is incorrect? How would you know? Hybrid intelligence is solving this issue. With the AI model trained and monitored by the people who know how all this is supposed to work, it is much more accurate than a generalist AI tool left to its own devices. Especially when the enterprise frequently updates customer processes or policies and agents need access to the very latest information. 

With examples like these, it is clear there are many new sources of value if you are willing to start building with AI. Yet overcoming enterprise inertia is still a major obstacle if your organisation is to be an early adopter rather than a laggard. With AI engineering experience and skills at a premium, enterprises will be looking for external support. From people who have already delivered production systems, who understand the challenges of getting them live and who know how to bring together humans and machines.

The trouble is that the traditional consultancy approach cannot deliver the early cognification required. While AI product vendors cannot deliver the hybrid intelligence that enterprises need to deliver real value and become early adopters in their industries.

As the cognification of industries gathers pace, there is a clear need for enterprises to identify where and how smarter processes can deliver the most value. Brightbeam was founded to tackle this problem and help enterprises gain advantage through AI by transforming how humans and machines collaborate.

The founders’ vision for Brightbeam is to work with early adopters to support cognification, unlock hybrid intelligence and deliver advantage through AI. 

Head of Strategy, Phil Black, describes how the company came into being, “We felt that enterprises were being short-changed by consultancies that could only talk around the issues. Instead of just saying what’s possible, there is a need to show it too. So rather than tinkering with proof of concepts, we stand up safe, working production systems that really prove the value before moving into more challenging areas. And starting anew in the Age of AI enables us to avoid the pitfalls of traditional consultancies by combining the best of a service company with the best of a product provider.”

Chief Operating Officer, Paul Savage, says that while there is lots of hype surrounding AI, this is also the best time for enterprises to get building with the right kind of partner. “Each of us at Brightbeam shares the same excitement about the potential of AI. We have all been involved in technology change and seen how it affects the enterprise. We’ve seen what can be gained from digital, big data and machine learning. We have all been involved in AI – in one form or another – for nearly thirty years. The biggest impact is when you build, deliver and learn quickly. This is where our experience in building and deploying enterprise software makes the difference. Enterprises don’t always have this combined experience in-house.”

Having seen first-hand the factors that have led to the success – and failure – of building within enterprise environments, the founders understood how simply jumping in and going it alone would not deliver the value AI promises. Firstly, in finding and, secondly, in quickly developing high-value use cases for hybrid intelligence within the individual enterprise context.

As CEO Brian Hanly explains, this is central to Brightbeam’s core mission: “We want to be the most helpful company in the world. So when advice is needed, we give it. When support and guidance are required, we consult. But our bias is for action and collaboration – building fast, learning quickly, and delivering value but always within a specific enterprise environment.”

Adapting to the Age of AI is still new to most people. And enterprises in particular need a special kind of support in kick-starting greater cognification. 

Simply jumping in and going it alone will not work. Those who do but then struggle with live data, bridging the gap between humans and machines or addressing security concerns will become the ‘cognification laggards’ who spend years trying to catch up. While those enterprises that quickly identify valuable opportunities for cognification, develop their own sources of hybrid intelligence and build early yet securely will be the ones that gain advantage through AI.

Brightbeam partners with the latter.