Time for Artificial Intelligence to Really Deliver

Technology is a sector that has often been prone of over-hyping the latest trends that emerge within it. Over the past few years, virtual reality and driveless cars are just two examples of technologies that while rich in potential, have been over-hyped by both tech and business press, when they are in fact some way still from emerging fully.

The fact that industry analyst group Gartner calls one of its market analyses the ‘Hype Cycle’ further highlights the tech industry’s propensity for hype, and perhaps one of the most over-hyped technologies of recent times is Artificial Intelligence (AI). Despite the media attention and the number of vendors claiming to have world-leading solutions, the number of actual end-users deploying AI to good effect is fairly limited.

But that’s not to suggest that AI is without enormous potential. A recent (July 2017) report from the McKinsey Global Institute Study, Artificial Intelligence, The Next Digital Frontier, revealed that tech giants including Baidu and Google spent between $20B to $30B on AI in 2016, with 90% of this spent on R&D and deployment. So AI is on the agenda of the biggest and smartest firms in the world. But how can organisations make the best use of AI and ensure it truly delivers on its potential?

Data and AI – the perfect match

The one thing that can help AI emerge from the hype is simple – data. Organisations hold more data than ever before, and thanks to the Internet of Things and the connected world we live and work, this data is growing all the time.

Used effectively, data can deliver unparalleled insight into customers and more, but it has been a challenge until now for organisations to manage this data in a way that enables easy extraction of actionable insight. The answer lies in AI and machine learning (ML).

But before AI and ML can get to work, a big issue for many organisations is the manner in which data is stored. It isn’t uncommon for a business to store and manage data in multiple CRM platforms, as well as a number of other repositories – ERP, other databases, on the server and desktops – across the enterprise.

This can arise through M&A activity, expansion or simply inefficient managing of IT resources, but the result is the same – data held in many siloes. This makes the management and analysis of this data a much harder job than it need be. If you cannot access data than how can you be expected to draw insight from it?

Addressing unstructured data

Furthermore, the sheer volume of big data in modern organisations can be bewildering, and it comes in files and formats that most CRM systems are unable to manage effectively. Unfortunately, this data is often the most valuable, containing rich insight into a particular customer and their specific needs and requirements.

This unstructured data includes: any social content – Twitter, Facebook, LinkedIn, Instagram – by, and relating to that customer; email conversations between the customer and brand; service call scripts that detail any recent or historical issues, and much more besides that doesn’t into the formats used by most CRM systems.

By not deploying unstructured data within a CRM, it can potentially be a major problem. It means that huge swathes of potential customer insight are missing. Utilising technology that captures both structured data in siloes and the masses of unstructured data, means businesses can begin to benefit from AI.

The potential of AI

Used properly, AI it can be hugely transformative and have a genuine and tangible impact on businesses and their customers. For instance, enterprise search is an area that is crying out for the use of AI and ML. Knowledge workers are spending too long searching for information that might not even be filed where they are searching for it.

Using an AI-based cognitive system allows much more complex search queries and can even make relevant and contextual suggestions back to the user. This is AI effectively understanding better than the user what that user is looking for, saving time on searches and delivering better results.

AI can also deepen customer understanding. An issue for larger firms is knowing who holds which relationships at a particular customer. AI technology can look at mases of unstructured data – all emails sent to an organisation, from and to different people, in different departments and multiple locations – and provide a way for a user to take meaningful action with a customer. This outlines clearly and in real-time who knows who within an organisation, invaluable when enterprise relationships can be so complex.

These are just two examples – the potential of AI can go much further than that. But the key to taking AI beyond the hype lies with data, that’s where the potential is at its richest. By deploying AI and ML, organisations can collect data from multiple sources and in multiple formats, extracting fresh and insightful meaning from it and helping to deliver a complete view of that customer.

First appeared in CompareTheCloud.

Posted in Artifical Intelligence, Internet | Comments Off

Top 100 Digital Shaper

Feeling humbled and honoured to have been selected to be among so many great people shaping Switzerland’s digital future! A great team is the source of this success! #DigitalShapers




Posted in Squirro | Comments Off

You can see the AI age everywhere but in the productivity statistics

“At the hospital they introduced a computer to do payroll. Instead of four people in the payroll department they now employ eight plus a few in the new computer department,” my father explained over Sunday lunch over 30 years ago.

My father worked at the time as medical doctor at the citizen’s hospital (Bürgerspital) in Solothurn, a midsize Swiss town. Back in the eighties the first wave of computers made their way into any large organization, quite often payroll was one of the first applications.

Payroll was a great problem for early computing: A relatively easy task – Take the salary number, deduct social security charges and issue a bank wire statement, highly repetitive – every month the same, and a volume play – the hospital employed hundreds of doctors, nurses, and auxiliary staff.

This was exactly what early computers where good at: straight forward repetitive tasks multiplied many times.  While the computers mastered their designated assignment well, the actual challenge was to embed this new technology into the daily routine at the hospital’s payroll department.

This was difficult for several reasons. For starters, computers were not flexible, so you had to adjust processes to the computer and not the other way around. The early computers were quite big boxes and error prone, especially the matrix printers of the time. Next to the new technology challenge you had a change management process on your hands. The answer quite often: You had the old and new processes run in parallel for a while to account for teething troubles.

Economist Robert Solow famously quipped: “You can see the computer age everywhere but in the productivity statistics.”

Some thirty years later we are at the dawn of a similar transformational phase: The advent of Artificial Intelligence, or AI in short. Every day someone claims another AI breakthrough. It’s the age of intelligent machines, automating everything and taking over the world, or so we’re told.

I beg to differ.

Instead I see a bit of history repeating itself here. Let me explain.

While Siri, Cortana, and similar services start to be great at simpler tasks, they stumble easily if asked more complex questions, or questions without context. Elsewhere you see early forms of machine learning and neural networks applied to repetitive tasks such as identifying cat pictures with limited results.

Let’s inject a dose of pragmatism: Currently it’s faster computing multiplied with clever algorithms.

And many of these algorithmic methods have been around for a while. For example, one of our core USPs, the catalyst detection is built on the shoulder of giants. The work of Robertson-Sparck-Jones. The main body of this work has been developed in the 1980 sand 1990s. We expanded the work on probabilistic retrieval and applied machine learning techniques in the period 2012-2014 together with folks from the University of Applied Science of Zurich and ever since.

Here at Squirro we deploy these techniques to do more with data. And some of the results are quite amazing. Have you seen a system not knowledgeable of what the Octoberfest is to accurately predict the next, using no more than a few time buckets of available data?

To quote Marc Vollenweider from Evalueserve: “We’ve tested 25 #AI engines & only @Squirro brought benefit.

Independent evaluation from the market gives rigor and context to such new technologies and helps to distinguish today’s useful and practical AI, from the hyperbole that surrounds it. We were not aware during the initial assessment phase at Evalueserve of their selection process. And obviously humbled and proud about it’s outcome.

The AI story has a long way to go and won’t make that journey unassisted. As Marc explains in his recent book ‘Mind + Machine’ the phase of combining the best of machines and human brains is where we can expect to see great value. I agree.

To a certain extent, it’s the same story from 30 years ago. The promise of AI is real.

Within a limited compound, some AI products already produce stunning results. A good example is the progress made of applying AI to language translation. But they are still limited in scope and limited to this box. The everyday embedding piece is often missing.

We’re living in an age of experimentation. Along the we learn of every stumble and hiccup and that’s a good and necessary thing for any new technology. Exposure of limitations de-mystifies it over time, making it more likely to be accepted, adopted and find its most useful and practical applications.


PS: Reach out to us, as we’ve gained considerable experience in deploying AI solutions providing tangible results.

PPS: Would I be an aspiring economist I’d explore the reapplication of Solow’s thesis.

PPPS: Image credits

Posted in DE, EN, Society, Squirro | Comments Off

Digital Manifesto for Switzerland

The other day Digital Switzerland invited 50 digital movers and shakers to Bern to discuss digitalization and Switzerland and what we think a way forward is. The outcome is this Digital Manifesto (pdf).

Posted in DE, EN | Comments Off

“We’ve tested 25 #AI engines & only @Squirro brought benefit”

As known, I work for Squirro. Our customer Evalueserve just released this short video:

“We’ve tested 25 #AI engines & only @Squirro brought benefit. This is why #robots won’t replace us”

@Evaluserve, 23 August 2016

I agree.

Posted in Curation, Squirro | Comments Off

New York New York, big city of dreams

New York Central Park

The big apple at 6am, just returning from my morning run in the Central Park. It’s already nearly 80 degrees Fahrenheit. And it was busy in Central Park, the center drive full of joggers and bickers.

I am here to join my team for a full week of customer, prospect and partner meetings. We’ve established our New York office last year and have with Peter a fantastic account executive and with Alex an amazing post-sales engineer on board.

Together with our team in Europe we work on a number of customers in the financial services world and beyond. And we broaden the reach daily. Just heading to a prospect together with our friends from Synpulse, our great consulting & integration partner here in the US.

The US is a great place to do business. The forward leaning approach, taking in new technologies to reap rewards early, helps a younger vendor such as us. The positive attitude to risk and risk taking (in considerate amounts) is a sharp difference from the risk resistive approach that I often see across Europe.

So over the next year or two we’ll expect to win quite a number of new customers & and add a few more colleagues to our team. And I expect to come many times more for an early morning moment to the Central Park.

Posted in Heute in | Comments Off

A year on

UntitledIt’s really the light. Coming into Charlotte two days ago there was this wonderful late afternoon light across the city, a kind of deep blue, combined with southern warmth. Later the sky turned warmer still. The reciprocal effect of light and clouds gave the skyline a mysterious and beautiful flavour.

Downtown Charlotte


I came here to Charlotte, North Carolina, the first time last year. We were visiting Wells Fargo, a US bank. Today, 12 months later, I am back to visit a visionary team, transforming the way the bank does more with data.

For the past decade any bank faced an ever-growing level of competition. As a reaction Wells Fargo puts long lasting client relationships in the center of its activities. With customer satisfaction being directly linked to competitiveness and profitability, the bank needs to leverage new technologies and do more with their (existing) data to create deeper and wider relationships.

The bank has chosen Squirro to deliver this vision. We’ll post in the near future a more complete customer success.

Posted in Banks, Heute in | Comments Off

Brexit: See EU later

The people in the United Kingdom voted for Brexit – My prediction was wrong. It’s just the opposite of what I expected in terms of percentage but correct in terms of the Scottish result.

Tough stuff.

As always we all will muddle through, but it’s unchartered territory.

At the risk of again being wrong about predictions: It’s possible that Cameron will go down as the guy who undid two unions.

Posted in Uncategorized | Comments Off

The Outcome of the Brexit Vote

Here’s my prediction for the Brexit outcome: It’s going to be ‘Remain’ by 51-52% overall, the English will vote out and the remain will only prevail because the Scots vote by about 60%+ for the remain side. The result will be increased tension within the United Kingdom.

Posted in Uncategorized | Comments Off

Make data human – Apply Deming’s principles to the industrialization of data analytics



It was a sunny spring day. I just received my coveted birthday present. A Sony Walkman. It was revolutionary. Small, lightweight, beautiful, it was music on the go. And it was the essence of Japanese ingenuity.

W. Edward Deming knew a few things about counting: He was the son of a chicken farmer in Iowa, USA. A trained mathematician, he worked as a census consultant to the post-war Japanese government.

While there, he was asked to hold a short seminar by the Japanese Union of Scientists and Engineers. He taught statistical process control and concepts of quality.

Deming called his system of thought “System of Profound Knowledge”. His message to Japan’s chief executives: Improving quality will reduce expenses while increasing productivity and market share.

Many more seminars followed, with one of the attendants being Akio Morita, the cofounder of Sony.


Industrialization of data analytics

Deming’s methods profoundly transformed the industrial processes in Japan. It’s time to apply these same concepts to data analytics.

74% of firms say they want to be “data-driven”, reports Forrester. Yet only 29% are actually successful at connecting analytics to action.

Rajeev Ronanki et al. of Deloitte Consulting pointed in a recent blog post to some of the reasons for this apparent contradiction. They outline:

“Advances in distributed data architecture, in-memory processing, machine learning, visualization, natural language processing, and cognitive analytics have unleashed powerful tools that can answer questions and identify valuable patterns and insights that would have seemed unimaginable only a few years ago.


Against this backdrop, it seems almost illogical that few companies are making the investments needed to harness data and analytics at scale. Where we should be seeing systemic capabilities, sustained programs, and focused innovation efforts, we see instead one-off studies, toe-in-the-water projects, and exploratory investments.”


Data Analytics Principles

It’s time to change and a good place to start are Deming’s methods. Deming advocated in his System of Profound Knowledge four key points:

  • Appreciation of a system: understanding the overall processes
  • Knowledge of variation: the range and causes of variation
  • Theory of knowledge: the concepts explaining knowledge and the limits of what can be known.
  • Knowledge of psychology: concepts of human nature.

Let’s apply these four points in turn to data analytics.

  • Appreciation of the system: any analytics initiative should be setup with the goal of improving products or services. This may include suppliers, producers, and customers (or recipients) of your goods and services. Any analytics initiative must have to goal to provide novel, timely and actionable insights in context, relevant to specific production process.
  • Knowledge of variation: Analytics today is correlation. Regardless of the level of sophistication any correlation has statistical sampling issues.
  • Theory of knowledge: Deming railed against blindly asserting opinion as fact, out of convenience or ignorance. At the start of any analytics initiative a company lacks the frame of reference to validate and assess results. A good way is exchange results between industry partners and providers (we’re ready to share ours) to learn what is necessary to improve the situation. Learning needs to be continual and organization-wide.
  • Knowledge of psychology: Deming understood the fundamental truth that people are different. Indeed one can create the best analytics system, know all about variation and still have a failing analytics initiative. The key is to understand people, and particularly what motivates them. The transformational effects of analytics are profound. The key to is make people not just part of such a journey but address intrinsic needs, including taking pride in workmanship and working with others to achieve common goals.

Example: One of our customers is rolling out our Service Insights solution. The key goal: optimize their call center response times by up to 30% (in fact deploying the pilot results across the entire call center). As part of the initial project setup we involved the call center agents in the actual design of the solution.

The effect: the agents were driving the project. It was no longer a management imposed efficiency initiative but a team effort to improve their workplace The team made use of data to transform their organization. In a way they made data human.


Image credit: drgautamgulati.com

Posted in DE, EN, Organizations, Squirro | Comments Off