“Do you have any experience with AI?” – this question has started cropping up regularly in the inquiries the Railsware team gets every day. We’re a product studio with a more than 10-years of experience in different focus areas, and artificial intelligence is one of them.

Today, AI can be applied anywhere from retail sales to a rocket launch. It attracts the smartest minds and causes the increase of both entrepreneurial activity and investment flows every day. In this article, we’ll give an overview of the London AI ecosystem and current development of AI in the UK.

Why London? This city is undisputedly one of the world’s most interesting tech capitals. We’ve conducted solid research about AI in London, and you’ll find no assumptions or baseless speculation. Our goal is to share our knowledge of this space and our expertise in AI to provide a better understanding of where the opportunities for startups lie and how you can grasp these opportunities.

What is AI and where is it implemented?

The era we’re living now can be rightfully called AI-powered: algorithms driven by machine learning, natural language processing, neural networks, image recognition and other subfields of artificial intelligence are leveraged in a tremendous number of applications from chatbots to self-driven vehicles.

The term artificial intelligence denotes the capability of any code or algorithm to imitate the cognitive process of a human being. So, an AI-powered machine can make its own decisions as for planning, learning, recommending, and understanding data. Below, you’ll discover the most known focus areas and implementation cases of AI.

The solutions built with machine learning, computer vision, and other incarnations of AI are conquering a variety of industries including finance, insurance, nuclear engineering, and so on. Meanwhile, we can evaluate the so-called AI revolution from the viewpoint of geographical expansion because the technology is being put into action not only across tech giants like Google or Tesla but also across countries. Traditionally, the United States and their technological antagonist, China, are the frontrunners.

However, there are several large economies that also compete for domination in AI. The United Kingdom is one of these ambitious contenders, having the capacity to bring innovation topside and accelerate AI adoption.

Why is London the artificial intelligence capital of Europe?

The third-most-populous continent on the globe could become a major hub for the AI boom. Within the next two years, Europe is expected to raise $24 billion investment from both public and private entities. Despite the fact that Europe comprises numerous developed countries, only a few of them are poised for an artificial intelligence breakthrough. These are Germany and France, and their capital cities have managed to raise over $110 million and $400 million of venture capital funding into AI companies for the last five years respectively.

But the UK is in a class of its own. This country has made great progress in AI development, and now it can boast much more significant investment into this sector and richer government support. In 2017, the United Kingdom was the fourth in the global AI race according to multiple studies. Today, it’s already the third and is on course to increase its ranking even further. And a beacon role is ordained to its capital city.

London, being one of the most populous cities in Europe, has the ambition to become a smarter place from the viewpoint of residents, tourists and entrepreneurs. For that purpose, the Mayor’s office has developed a dedicated commitment, which rests upon the integration of AI-driven technologies into a wide range of life areas including healthcare, insurance, communication, security, etc. This goal is not achievable without the involvement of the best minds, investors, and entrepreneurs, which in turn, require fertile ground for growth. London is covering all facets of data science implementation at a swift rate and turning into a place you need to be in for the research and development of AI.

AI ecosystem in London counts 750 data science startups

The UK’s capital keeps enlarging the AI ecosystem, which speaks for a favorable environment that professional engineers, entrepreneurs and startup founders can discover in London. Today, the city has already become home to over 600 companies that sell at least one AI product. The total number of data science startups London exceeds 750, among which there are representatives of different industries including marketing (MiQ), entertainment (Improbable), healthcare (BenevolentAI), advertising (Iponweb), and others. Meanwhile, the international standing of the city rests on the industries of law, insurance, and finance, which in turn, are auspicious for rolling out a strong AI ecosystem.

According to the number of AI companies, the UK’s capital city outstrips its European competitors, Berlin and Paris, and ranks close to the San Francisco Bay Area, which is the global largest AI ecosystem.

On a national scale, London remains the center of talent, innovation, and customers. Most British AI companies choose this city as their home. Numerically, 40 of 50 top AI startups in the UK settle in the capital.

Investments in AI startups hosted in London exceed $400 million

The AI London startups are not short of investments, which stand at massive amounts. Since 2013, artificial intelligence companies UK got around $1.61 billion of financial placements, and the capital’s share is over 65%. In 2017, the VC funding of Berlin and Paris was $22.42 million and $126,16 million correspondingly. Even taken together they cannot surpass London’s investment, which is more than $250 million. The money flow in 2018 exceeded $400 million with only BenevolentAI managed to raise $115 million in April. All these figures prove the London’s leading investment appeal as for artificial intelligence in Europe.

Despite the UK’s third rank among the largest AI destinations in the world, London lags behind cities from China and the USA as for the VC funding volume. San Francisco keeps retaining the leading position in terms of AI investment. Though it failed to be ranked first in 2017 and gave way to Beijing and Shanghai with their $2.07 billion and $2.03 billion respectively, Bay Area returned to the throne in 2018 with over $2.3 billion attracted. New York became the second AI investment-attractive city with its $1.5 billion.

The UK capital is one of the most attractive destinations for AI talent

The talent base is a fundamental component of the long-run development of the ecosystem. In this regard, London is one of the most attractive cities around the globe for AI functional specialists like machine learning engineers or data scientists. The UK capital is surrounded by talent foundries such as Oxford and Cambridge. Besides, many AI experts recognize London as a favorable location in terms of tolerance and diversity.

The current state of affairs shows that the UK ranks second in the global talent pool with around 1.8K high-profile candidates available on LinkedIn. A lack of formal policies conducive to talent retention led to brain drain in recent years. Now, the tables have turned, and both the UK government and London mayor’s office are focusing on growing the city’s talent base and attracting a world-class supply of AI expertise.

London is investing in AI ethics expertise

Having a strong commitment to gain the lead in the AI development, the UK is focused on making London a center for the ethical development and deployment of artificial intelligence systems. The major goal of the initiatives coming from both the government and the third sector (non-governmental and nonprofit organizations) is to ensure the public understands the AI benefits and is aware of challenges associated with its misuse. This will have a positive footprint on the economy of London and the entire UK.

The series of actions to be taken for the growth prospects of AI ethics include the creation of a Centre for Data Ethics and Innovation to the tune of $11 million. It will analyze how AI is governed and provide recommendations to the government thereto. The role of public events dedicated to AI should not be undervalued. Such famous conferences as Silicon Roundabout, AI & Big Data Expo Global, O’Reilly AI Conference, as well as The AI Summit London, will pave the way for engineers, product managers, entrepreneurs, and executives to make difficult strategic decisions associated with AI’s impact on the business.

AI adoption in London’s industries

Despite a rather low percentage (20%) of companies using artificial intelligence technologies in the leading industries of London so far, the adoption of AI is increasing since the interest in innovations is growing. Most startups prefer to target other businesses (B2B) rather than final customers (B2C). London has a large number of AI suppliers in global industries such as education, finance, healthcare, insurance, law, media and entertainment, retail and sales, and marketing. These industries represent huge future growth potential for London’s AI ecosystem.

Education

The UK ranks first in venture capital funding of educational technologies in Europe. This industry is expected to reach $4.3 billion by 2021 and has a large potential market for AI development. Nevertheless, the current adoption rate of Edtech is rather low due to the lack of funding and insufficient involvement of teachers in product development. Meanwhile, low adoption is not a blockage, and numerous startups are already engaged in providing AI-driven solutions for both students and educational establishments. The examples include Currikula, which leverages ML algorithms for checking plagiarism and analyzing assignments, Century Tech, which uses the benefits of AI for personalized learning, and others.

Railsware’s experience in the education industry is represented by BrightBytes, an online data analytics platform, which is meant for analyzing students’ performance and predicting their dropout. Machine learning lies at the heart of BrightBytes prediction model. We were involved in the most challenging part of this project – building the data set to be used by ML algorithms. Our team of engineers proficient in Ruby on Rails and other technologies together with talented educators who shared their deep experience in K-12 educational system analyzed piles of data to filter out the otiose ones. From our side, we’ve leveraged our best engineering practices in making some key decisions to help BrightBytes grow to the level of a globally known solution in Edtech.

Finance

London, being one of the leading financial centers of the globe, has all prerequisites for nurturing a healthy FinTech ecosystem. Today, it counts 140 AI-driven companies engaged in the financial industry. The adoption rate of machine learning technologies in finance is higher than in Edtech but varies depending on the use case. For example, retail banking falls short of AI technologies, while algorithmic trading or financial management have enough solutions available on the market.

The most prominent representatives of AI-driven fintech startups in London are Cube (regulatory change manager), BMLL (ML-powered analytics platform), and Ravelin (fraud detection and risk management solution). Other cases where AI can bring value for financial services include credit and payment data analysis, algorithmic trading, customer engagement, and market analytics. Here you can learn more about machine learning in the financial industry.

One of the most notable examples of Railsware involvement in London fintech software development projects is the collaboration with Zephyrus Partners, a strategic pensions advising firm. This client felt the need in automating the PDF data extraction process. Instead of machine learning software development services, the Railsware team applied the concept of artificial artificial intelligence (AAI) to perform a huge amount of small tasks. We also used Google Visual API for file parsing and data extraction and Amazon Mechanical Turk for data verification. As a result, we built a quick prototype of an app and provided the client with a solution to cope with over 3k of PDF files. You can read more about this AI London use case here.

Healthcare

This industry is the most active for investments focused on artificial intelligence technologies. Globally, the health AI market experiences a huge growth with the prediction to reach $6.6 billion by 2021. The use cases typical for machine learning implementation in the medical sector include diagnostics, personalized treatment, and administrative processes. Nevertheless, the adoption rate in London is quite low, and the number of AI startups is around 60, which is almost 2.5 times less than their number in the Bay Area. The most renowned ones are BenevolentAI and Babylon Health. The former offers AI-powered software to target identification, clinical mechanistic stratification, and molecular design. Babylon Health leverages machine learning and natural language processing for efficient reasoning, transcribing consultations, summarising clinical records and other healthcare-related activities. Other London-based startups include DeepMind, HealthUnlocked, Medopad, Your MD, Kheiron Medical, and so on.

Insurance

Insurance is considered one of the most attractive industries for data science. Many experts predict that AI will bring significant changes to InsurTech like automated underwriting for small businesses and private entities, as well as a significant headcount reduction in the next ten years. The potential impact of AI on the industry is estimated at over $200 billion. At the same time, the current adoption rate is far from ideal. Compared to the US, where this rate is around 40%, the entire UK boats just 9% AI implementation in the industry.

Nevertheless, London is pushing progress, and, in 2017, the InsurTech investments grew by 19 times and reached $364 million compared to the previous year. Technologies are leveraged to improve underwriting accuracy, claims management, marketing, and customer experience, as well as to detect fraudulent activities. The most notable AI-driven companies on the market are Artificial Labs (offers versatile InsurTech tools for predictive analytics or customer interaction), Cytora (risk management), Quantemplate (a bunch of products for claims efficiency, underwriting capacity, etc.), and others.

Law

Total revenue from legal activities in the UK in 2017 was $42.2 billion. As of 2018, the number of UK law firms using artificial intelligence exceeds 20. The activities optimized with data science techniques include document generation and review, due diligence and versatile researches. The AI revolution will rehash the industry, and junior lawyers and apprenticeship system will experience its impact at most. London has made a breakthrough in legal innovation. Nevertheless, the adoption of artificial intelligence in this industry will have to overcome many challenges like insufficient AI productivity at early stages of implementation in the law firms, as well as the upcoming Brexit, which is likely to have a significant footprint on the law-related activities.

The UK’s capital has a solid pool of AI startups to strengthen the modernization of the legal sector. Such companies as Eigen Technologies, Luminance, and Juro have already established a successful partnership with law firms in London and abroad. The use cases of AI for the legal sector include document processing and automation, research, due diligence, legal analytics, billing, and litigation outcome prediction.

Media and entertainment

By 2021, the total revenue of media and entertainment is expected to exceed $90 billion in the UK alone. This sector is vast and one of the most influenceable from the viewpoint of digitalization. London has a considerable AI startup base of over 70 product/service suppliers. Data science technologies are mostly leveraged for visual effects, content creation, personalized UX, media manipulation and retrieval, as well as gaming.

The rate of technological adoption is significant due to inexhaustible investment opportunities. For example, a renowned London-based startup specialized in building virtual worlds – Improbable – managed to raise half a billion USD in 2017. Other notable companies include Foundry (design of creative visual experience), Lobster (image recognition for search), Jukedeck (machine learning-based tools for creating music), and so on.

Retail and e-commerce

In 2019, the industry of retail and e-commerce is expected to approach $30 trillion on a global basis. Along with that, the growth of artificial intelligence will lead to massive changes mostly related to the automation of processes on all levels of the retail chain. In the UK, around half of all retail companies recognize themselves as AI-driven organizations meaning they use neural networks, machine learning, computer vision and other AI techniques for customer recommendation engines, sales forecasting, fraud detection, automation of mundane logistics and delivery tasks, virtual assistants, and visual search.

The most notable representatives of London AI startups are Hoxton Analytics, which is known for its IoT edge sensor using computer vision and machine learning algorithms for counting footfall, Snap Tech, an AI-powered visual search platform, Cortexica, which offers technologies for image recognition, video analysis and search, and many others.

Marketing and sales

The last but not least industry in our pick is predicted to have a $2.6 trillion impact on the global economy in the future. Marketing and sales are experiencing gigantic transformations because of the growth of artificial intelligence. The most evident outcomes are the personalization of messaging and programmatic purchasing. Nevertheless, the AI adoption rate in the UK and London, in particular, is still in a nascent condition.

The major obstacles in the way of a widespread AI implementation are the lack of funding and understanding of how new marketing technologies can bring the production to a progressive and efficient level. As a result, London’s AI adoption for sales and marketing lags behind San Francisco’s by almost three times. This industry requires an impulse, and such companies as Codec (an ML-driven platform to discover and track relevant cultural networks), Qubit (data analytics tool for personalization, product discovery and recommendation, etc.), Iponweb (infrastructure and technology provider for online advertising platforms), and others have a potential to increase the presence of data science technologies in marketing and sales.

In general, the adoption rate of artificial intelligence in London ecosystem is in its early days. The exception covers some companies and industries where AI-driven solutions have been already put on the job. The common use cases include algorithmic trading, image and voice analysis, and recommender systems. Nevertheless, we can observe a significant rise in demand, as well as interest in innovative products. London-based companies are eagerly applying themselves to machine and deep learning and becoming more sophisticated AI adopters. This positive curve will kickstart the enhancement of the AI startup base and lead london to the top of the AI ecosystem worldwide.

The real reasons why startups need AI

Almost any newbie startup is willing to augment its name with two magic capital letters – A and I. In most cases, that would just make a product sound cutting edge. Many successful AI startups confess that their major fundraising hook was “we’re AI-based”. For the reason that the term is rather indistinct, many entrepreneurs and companies are taking unfair advantage of using it in relation to their products or services.

There are numerous pitch decks stuffed with references to “artificial intelligence” and “machine learning” to simply raise the amount and likelihood of funding. As a result, you can encounter some “kind of” AI products or services with simple heuristic methods or database processing techniques under the hood. And here is a typical example of how it works:

The company X markets itself as an AI platform meant for the improvement of their clients’ financial health. It is an expert system that can give suggestions regarding operations management in transportation, retail, hotels, and restaurants.

With that in mind, they do not leverage any machine learning algorithms but rest on statistical methods to build their recommendation engine. The entire model is based on a decision tree, according to which the clients get forecasts for their business activities.

The difference between true and unfair AI power lies in the availability of ML-based algorithms for building either a forecasting system or recommendation engine or anything else. The true AI means that you do not have to say a machine how to achieve the goal and the machine should learn it by itself using the input data.

Pick one of the following strategies for your AI-driven venture:

Apart of any show-off intentions to become more sticky for investors, there are only two major reasons why startups crave to get a hold of the AI power.

1. Winner Takes All

One of the greatest promises the AI is associated with is a higher probability for a startup to be acquired within a relatively short period. In most cases, it is 3-5 years. Artificial intelligence can be a significant advantage over the competitors within the industry, and the market entering perspectives seem bloomy when your product leverages machine learning, neural networks, natural language processing or another data science technique. As a rule, the winner-takes-all concept is applicable to companies with access to proprietary data. In the future, they can improve their product by acquiring and using this data, and stay ahead of the competitors.

You can discover a bunch of examples of successful London-based projects that managed to get acquired rather fast. Firedrop is three years old and is already a new category of automatic website builders, which delivers creative vision at scale with machine learning. LiveJump, an energy company, which provides aggregation, power purchase agreements, trading and supply services through big data, was founded in 2013 and now has more than $5.5 million of total funding. One of the MedTech giants, BenevolentAI, was also founded in 2013 and now exceeds $200 million in total funding amount.

So, AI gives an unfair advantage for startups to enter the market and set dominance in the future.

2. First of a kind

Another reason that makes AI attractive is its capability to disrupt sectors (finance, insurance, healthcare, and others) that are not taken by big tech companies. Many founders abandon the direct competition in their long-term vision and prefer to deliver something first of a kind to the market. And artificial intelligence is meant to help them get this through.

A disruptive startup must bring to life groundbreaking ideas and ambitious plans. DeepMind is a good example of a disruptive company in healthcare. In addition to the leverage of AI technologies for radical improvement of healthcare, the company is engaged in dealing with other global issues like climate change and employs different scientific advances to find solutions.

Other remarkable London-based companies that we can call disruptive include Headstart, a ML-based recruiting platform, Sentient Machines, a tech company aimed at empowering smarter human communication using  deep learning, Drone Space AI, a software platform that enables complete autonomous operations in a swarm of unmanned aerial and ground vehicles, Proportunity, a machine learning-driven real estate lender, and many others.

In summary, a real AI startup pursues a goal to create a product where artificial intelligence is a constituent element of the value proposition to be delivered to the final user. Such companies do not use machine learning or computer vision for their own sake. They take advantage of AI achievements to solve an important customer problem and, by means of it, make money. If you are not sure about what value your product can offer to your customers, we recommend you take advantage of the value proposition canvas*** that we’ve explained previously in our blog post. This tool will help you understand the validity of your product and provide the rationalization for using the AI power.

How hard is the monetization journey for AI startups?

AI startups have better fundraising capabilities than traditional SaaS startups: as a rule, the cash raised at the Series A funding is up to 60% larger. At the same time, AI products incur significant rollout costs and are associated with an extremely high cash burn rate and increased talent acquisition costs. As a result, just breaking even can be hard to achieve. However, there is not a simple unique monetization strategy for AI startups, and the journey to ROI can be both short and long.

A long way to monetization

Truth be told, everything depends on your choice as a product owner. If your goal is to solve a specific problem using AI-based techniques, the road to monetization will be long enough. The reasons are that it is hard for AI companies to develop a robust Minimum Viable Product (MVP), the sales cycles are long in many industries for B2B products, and the deployment period for SaaS companies can be extremely long. Moreover, it is a challenge to sell a long-term project because it takes more time to deliver its value proposition. The examples of London AI startups that expect a long journey to monetization are Prognostic (a cognitive diagnostics & decisions engine), Multiply (a fully-automated independent financial advice service), Predina Tech (a prediction platform for self-driving vehicles), and others.

A short way to monetization

The way to ROI may be shorter if your company is research-driven or a founded startup is aimed at developing a new AI technology. In this case, you probably have to burden more investments at first. High rollout costs should be expected on attracting talents, developing infrastructures, processing data, etc. Nevertheless, the likelihood to capitalize even without generating revenues is higher. DeepMind is a classic example of a startup with a shorter way to monetization. Other examples of those based in London include Magic Pony (a company recently bought out by Twitter, delivering miscellaneous visual processing technologies based on machine learning and computer vision), AI4 THINGS (a company developing deep learning-based solutions for industrial robotics, delivery services, agriculture and pest control), Fifty (a technology company focused on analysis of social media data), etc.

To sum up, the performance of machine learning or natural language processing techniques is not characterized by a binary outcome (work or not) but rests upon the principle of improving with more data. It means, the bottom line of your AI product will depend on how fast it will be able to provide value, which, in turn, consists of building an AI model and acquiring or creating data for it. Once you choose the monetization way, it is worth thinking twice about the venue for rolling out your startup activities. And the British Isles have many benefits to attract entrepreneurs.

The UK is committed to helping startups succeed in AI

In 2017, the United Kingdom shaped a course of building a globally competitive data-driven economy. The aftershock of that was the announcement of the governmental commitment to AI known as AI Sector Deal. This paper outlines the key activities and measures to be undertaken towards the promotion of the domestic adoption and employment of artificial intelligence. These funds the total investment of over $2 billion for implementing will be distributed across the following five foundations.

• Ideas
  To become the driving force of artificial intelligence in Europe and further on globally, the UK will insistently invest in R&D sector to create new AI-driven solutions for different industries including law, insurance, power economy, healthcare, and so on.
• People
  In addition to fund-raising in specific academic development like digital and technical studies, as well as maths, there is an objective of establishing a competitive educational system to prepare talents for future in-demand jobs in machine learning, natural language processing, etc.
• Places
  The main goal here is to let the existing AI ecosystem thrive not only in the heart of Britain but also in other cities including Cambridge, Edinburgh, Manchester, and so on. The government is aimed at stimulating AI adoption among businesses and empowering ambitious tech entrepreneurs to grow faster.
• Business environment
  Entrepreneurs need open air to streamline their ideas, and the UK should become the best place for startups to blossom. The optimization actions include the establishment of an AI Council to deal with all activities around artificial intelligence in businesses, as well as the development of an adequate supporting policy.
• Infrastructure
  It’s one thing to build a data-driven infrastructure, and it’s totally another to make sure that it corresponds to the up-to-date capabilities like the last generation mobile networks, the availability of data sharing frameworks, etc. Therefore, the major focus of the infrastructure-oriented commitment is to keep pace with the times as for the citizens’ access to data-driven services and products.

To unlock their AI entire potential, the UK needs to maintain balance between business, academia, and government. This country is fertile soil for AI startups with such notable names as DeepMind or Babylon. However, the ambition of Britain and London, in particular, spread beyond the borders of the European Union. With the latest steps taken by the government, they have capabilities to take the lead in the AI revolution.

*** Value Proposition Canvas is designed by Alex Osterwalder, Strategyzer.com.