Artificial intelligence and machine learning solutions provider deepsense.ai today laid out its five-year strategy. Building on its accomplishments to date, the company aims to move further into the skyrocketing AI solutions market and grow fivefold in size by focusing its services on three key technologies: Predictive Modeling, Computer Vision and Natural Language Processing. It will pair that work with comprehensive Data & Analytics Platforms advisory and artificial intelligence/machine learning workshops.

Earlier this year, deepsense.ai was spun off from CodiLime to independently focus on artificial intelligence/machine learning projects and solutions for industry.
According to Tomasz Kułakowski, CEO of deepsense.ai, “a clear strategy and strong vision are crucial in building a successful IT business, especially in a cutting-edge segment like AI. By giving deepsense.ai independence to grow, we enabled the company to forge its own strategy”.

In the vanguard of progress

Analysts at Gartner and Deloitte believe the artificial intelligence & machine learning market is set to grow at a roughly ~50% compound annual rate over the next few years. “The opportunity is immense” says Kułakowski. “It’s no longer only a field for academic research. Companies are joining the party and want to see real business results from AI. Our full attention and a coherent plan on where we believe change will happen, and where we will focus our activities, were therefore essential”.

In addition to Predictive Modelling, Computer Vision and Natural Language Processing, deepsense.ai intends to invest heavily in advanced R&D activities, particularly in reinforcement learning, general adversarial networks (GANs) and environment simulation. This combination will keep it at the forefront of the AI/ML market. The company is already partnered with tech giants Google, Intel and NVidia, with whom it plans to further cooperate in delivering value to clients and continue joint research work.
“We see reinforcement learning as a technology of the future and that’s why we focus on it in our research. It can reshape the landscape in various ways. For instance, we are currently applying it in our work on autonomous vehicles” says Kułakowski. “There’s no challenge in following the trail blazers. Real technological leadership means being the trail-blazer”.

deepsense.ai will continue its development strategy by hiring talented data scientists and providing them opportunities to grow. “Business is always built by people, so hiring top talent is a cornerstone of deepsense.ai’s success,” says Robert Bogucki, deepsense.ai Chief Technology Officer. “We are and will continue to be an employer of first choice for data scientists.”

The last pillar of the strategy is built around providing tailored training to companies seeking to build AI competences.
“Our most formidable competition is not other AI providers, but doubt and a lack of practical knowledge about AI solutions. That’s why we share our knowledge with other companies during seminars, training and tech-talks, providing both education and mentoring for organizations that would like to transform their business using AI,” adds Kułakowski.

For more on deepsense.ai’s business strategy, see the presentation.

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate

CodiLime ranks 49th on the FT 1000: the complete list of Europe’s 1000 fastest growing companies. The company outperformed more than 950 companies from throughout the continent.

Technology dominated this year’s edition of the FT 1000: the complete list of Europe’s fastest growing companies, with 155 pure technological players. CodiLime was listed 49th on the strength of 2 057% revenue growth and tenfold employee growth in the 2013-2016 period.

“As a future–oriented company operating in the most innovative technological areas, such as cybersecurity, artificial intelligence and networks, I consider that growth a success that wouldn’t have been possible without the outstanding people we have on board,” said Tomasz Kulakowski, CEO and founder of CodiLime. “We are thrilled to be recognized by Financial Times among the fastest growing companies – this is the best proof that the market appreciates and needs what we build to help others live and work better.”

Ride the lightning

Europe has recovered from the recent financial crisis, and Britain’s looming exit from the EU has not proven as disastrous as feared. The countries of the monetary union significantly outperformed the US economy (2.3% expansion rate) with combined economic growth of 2,5%.

“In times of growth, companies look to invest in innovative business solutions. To provide our clients the best support they need to win on their markets, we offer services based on the most cutting edge technologies,” Kulakowski commented.

CodiLime delivers services in three key areas – artificial intelligence, cybersecurity and networks, with the highest focus currently on developing machine learning and deep learning models for both public and private sector clients. According to the Accenture Institute for High Performance, AI technology is predicted to double the economic output of developed countries by 2035. Today, only 15% of companies use AI, but 31% have said it is on their agenda for the next 12 months. AI’s market size is predicted to reach $59,748 billion by 2025.

This year CodiLime plans to significantly grow its networks offer and provide more services based on software–defined networking (SDN) solutions.

“As the world is more and more dependent on internet and seamless data transmission, we see high potential for growth in the telecommunications and networking sector. That’s why I consider the coming few years crucial to our further development in SDN”, Kulakowski said.

Facebook for whales and fraud detection

AI solutions, however, will remain a high priority for the company as there’s still a lot to be done. deepsense.ai, CodiLime’s AI–focused wing, was widely recognized for developing a deep learning model that could recognize right whales in photographs taken during aerial surveys. As the whales are on the brink of extinction, environmentalists and scientists from American NOAA had struggled to track the population in their effort to keep it healthy and safe. Data scientists from deepsense.ai built a facial recognition algorithm that could recognize a specific whale from a given photo. Their model achieved 87% accuracy, which was far higher than the human eye is capable of, and saved biologists the time, effort and guesswork involved with identifying the whales with their own, less exacting eyes.

“AI is presenting the world with countless opportunities, including fraud detection in the financial sector or customer churn prediction in retail. Some of them, like our “Facebook for whales”, come completely out of the blue. But this shows that the only limitation we have when creating new innovative solutions is a lack of creativity, so keep thinking outside the box!”, Kulakowski concluded.

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate

The National Science Centre in Poland has granted deepsense.ai data scientist and University of Warsaw researcher, Piotr Miłoś, 500,000 USD. The project “Reinforcement Learning – contemporary challenges” is designed to help resolve long-standing open problems in training robots and game agents in difficult environments.

The research is a continuation of deepsense.ai’s work on reinforcement learning, which was recognized by the General Chairs of the Conference on Robot Learning (CoRL) as one of 11 papers in the reinforcement learning and robotics category. In that project, members of deepsense.ai’s machine learning team trained a robotic arm to grip a can of coke using reinforcement learning. Recognizing that numerous fundamental ideas have yet to be discovered and understood, the company intends to invest in research and develop its reinforcement learning competences.

The grant, SONATA BIS, is extended to scientists with a strong academic background seeking to establish a new scientific group. Miłoś’ team will comprise students and postdocs from the University of Warsaw, and will be supported by deepsense.ai. Together they will address issues including hierarchical learning, curriculum learning, meta learning, metamemory, model-free and model-based methods.

According to Miłoś, “these are believed to be important concepts in developing reinforcement learning. One of the challenges in AI is the ability to plan hierarchically. deepsense.ai’s intention is to establish a group at the University of Warsaw that is engaged in world-class reinforcement learning research and will bring significant advancement in the field.”

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate

deepsense.ai’s parent company, CodiLime, has built a dedicated unit to provide organizations with comprehensive development programs to build their teams’ strategic competencies, starting from data science. The Hub also addresses managers’ needs in harnessing innovation.

CodiLime announced the Training & Development Hub at the company’s annual seminar on February 23. The new unit will support organizations in building competitive solutions driven by cutting-edge achievements in AI, cybersecurity and networks, three areas in which CodiLime has established strong expertise.

The Hub will offer long-term, customized development programs designed to grow team skills. Formalizing its approach to teaching, CodiLime has worked out and successfully tested its 4T methodology. The four Ts stand for TAILORED TEAM TRAINING TRACKS.

Marta Piechowiak, Director of the Training & Development Hub, explains, “Our decision to create the Hub was based on over three years’ experience in running workshops. These included open training for developers and data scientists, our corporate offer for global enterprises and successful execution of Intel’s strategy for its AI Academy in Europe. Our vision is to become a first-choice training and development partner for global companies seeking to expand their own expert teams in-house”.

Piechowiak adds, “Each of our training programs starts with an intensive few-day workshop which includes a theory section, followed by hands-on coding exercises and best practices. To make it even more effective, teams can benefit from regular mentoring sessions and develop their real-life projects under the guidance of an expert.”

The Hub will also support managers in their journey towards innovation. During the workshop, they learn how applying the latest technical breakthroughs can help their company win. The agenda also covers a brainstorming session and use case analysis for managers’ specific challenges and goals. This will prepare them to make better-informed decisions and meet their current and strategic goals.

The Training & Development Hub will be expanding its AI-based tracks throughout 2018 and launching new tracks in cybersecurity and networks in the second half of the year. The company has also started working on a collaborative platform for teams to make their learning experience even easier and more personalized, especially when working with a mentor.

The Director of the Hub adds, “Our data science training offer has been launched under the deepsense.ai brand. Companies seeking a partner to build their data science skills will discover tracks for reskilling software engineering teams, upskilling existing data science teams (with the newest program for advanced deep learning), as well as AI applications to help managers expand the ways they can incorporate AI into their business strategy.”

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate

Between now and the end of 2017, deepsense.ai will run 30 machine learning and deep learning seminars at prestigious universities across Europe as part of the effort to scale the Intel® Nervana™ AI Academy for students in EMEA.

deepsense.ai will train hundreds of European students in the upcoming three months as part of the Intel® Nervana™ AI Academy. Students will get the valuable opportunity to gain practical knowledge in two of the most cutting‑edge and fast developing areas of data science, machine learning and deep learning.

deepsense.ai organizes seminars and delivers machine learning and deep learning sessions, giving students a great chance to meet professional data science practitioners and ask them even the most challenging questions on advanced analytics. At the seminars, participants are also encouraged to join Intel’s Student Ambassador Program and work on their machine learning projects with infrastructural and financial support from Intel.

deepsense.ai CEO Tomasz Kułakowski stressed the importance of the collaboration with Intel: “I’m glad that deepsense.ai’s artificial intelligence expertise has been recognized by Intel and our company has been invited to support the Intel Nervana AI Academy in EMEA – that’s a project with great potential to successfully promote the most innovative approaches to data analytics today and grow new generations of talented data scientists from the leading academic communities. It requires time and full determination to build data science competencies at the highest level of excellence. deepsense.ai offers professional technical and business workshops as a part of our comprehensive approach to our clients’ needs – we’re open to sharing our know‑how with any company eager to upskill in AI.”

deepsense.ai shares Intel’ vision for making artificial intelligence solutions more approachable and widely used in daily life. To that end: deepsense.ai delivers break‑through end‑to‑end data science solutions while Intel builds the powerful hardware needed to process data effectively. Their experience in data science has allowed each of them also to create dedicated tools for data scientists, deepsense.ai Neptune and Intel® Nervana.

###

About deepsense.ai:

Media contact:

deepsense.ai trademarks at boilerplate