ING spins out Katana to become a fintech — Transforming an Advanced Analytics product to a company
ING announced the spin out of WB Advanced Analytics portfolio tool Katana to become an independent fintech company. What does it take to transform a product into a company? Frank Derks — Global Tribe Lead of ING WB Advanced Analytics and Santiago Braje — CEO of Katana Labs provide the answers.
Today ING announced the spin out of Katana, a cutting-edge Advanced Analytics tool that helps traders and investors make faster and sharper decisions when trading bonds. This is an important milestone for ING WB Advanced Analytics (WBAA) and Katana Labs proving that WBAA can enable the creative journey of a high performing team all the way from an Artificial Intelligence (AI) idea to an independently funded fintech.
But what does it take to make this happen? How does ING leverage Advanced Analytics capabilities to create successful products and companies that can benefit both ING and the market? We interviewed Frank Derks — Global Tribe Lead ING WB Advanced Analytics (WBAA) and Santiago Braje — CEO Katana Labs to find out.
Katana was conceived in December 2017, within ING’s WBAA tribe, and developed in an iterative process with FM Global Credit Trading team in London. The idea was to leverage the power of Artificial Intelligence and predictive analytics to enhance traders and investors decision making process. Identifying the need we wanted to complement their decisions with augmented intelligence without restricting their natural intuition. The users’ results reaffirmed our expectations, and Katana has been awarded as “Innovator 2019” by the Global Finance magazine.
Frank explains what sparked the idea to create Katana:
“WBAA’s hypothesis is that our clients are experiencing similar problems to ING so we can often empower them with AI products initially developed for internal use. Focusing internally first enables our teams to develop the knowledge, technology and science required before exposing this to our clients. So after creating a product for our own trading teams, we set up another team in parallel that ran a product design process with a number of core clients building on what we had learnt and developed internally. This worked out well, and after a number of iterations, experiments and pivots we created Katana labs”
Today Katana Labs has become an independent company, and we asked Santiago to describe the product’s breakthrough capabilities:
“Katana is a trade idea discovery tool for portfolio managers. We created it to scan the market systematically and detect anomalies in the relative value of bonds using advanced analytics. Our value proposition is to enhance portfolio returns by helping investment teams find profitable ideas faster and discover opportunities they would otherwise miss. Katana uses computational power and machine learning algorithms to analyse millions of trade ideas within an unbiased framework and help investors find the signals in the noise. There are no competing products in the market”
Katana was incubated within WBAA from a multidisciplinary squad of data scientists, software engineers, data engineers and UX designers. Enabling Katana to go through all stages of development and ultimately spin-out is challenging. Frank explains:
“Creating a minimum viable AI product and turning it into a minimum viable company, from within a corporate requires a lot:
First, You need to ensure that you have the capabilities in place to responsibly build the product and ensure the product works. You need a data analytics platform on which data scientists can work, compliance & legal ensuring proper use of data (although less relevant for Katana as we use publicly available data) and model risk management ensuring the quality of model output.
Second, You need a multi-skilled product team that brings user experience research & design, front end development, backend development, data science and engineering, product management and architecture so that the team can rapidly validate the problem to solve, and assess viability, desirability and feasibility of the product. Mandating the team to self steer stimulates high performance.
Third, You need stamina and a “get things done” attitude in order to deal with the challenges that arise in the product development process and in order to deal with complex questions from a variety of different stakeholders”
As ING transforms into a data native bank, technical infrastructure and expertise are indispensable, but culture also plays a major role in order to deliver disruptive innovations. At WBAA we have developed a unique culture that helps our people and product teams work in the most efficient way while enjoying and learning in the process. Santiago explains:
“ING WBAA is where the product incubated. From the very first conception of the idea, WBAA has provided the environment, technical knowledge and above all the culture to allow the idea to flourish. Katana would simply not exist without WBAA. As we approached the spin out in particular, the collaboration and focus on getting things done have made all the difference, helping the team overcome obstacles and accurately identifying the requirements at every step in the process”
During the Katana process we realized important benefits and learning both for ING and our clients.
Working intensively with core clients on a new topic has deepened the relationship between ING and these clients making it easier and more logical to cooperate and advise with ING more broadly. Client portfolio managers, have learnt what it takes to develop an AI product first hand while being able to influence its design in order to serve their needs. The Katana team has over time delivered trade insights to clients and ING traders that have been used for profitable trades.
Financial Markets management, traders, and the Katana product team have gained substantial knowledge about credit markets, relevant advanced analytics methodologies, our clients’ context and people, open source technologies, and market data; a key learning was the importance of design lead product development process.
WBAA ingested, structured and made available for use relevant credit markets and client data. We built a scalable (open source) architecture that we can reuse to the support the development of products that require massive amount of streaming data, fast response time and critical availability.
Frank also highlighted the transformational impact of Katana’s spinout:
“The experience gained by ING Financial Markets management and people through the process contributed to the transformation of ING WB into a data native organisation as people become more comfortable with the Data and AI concepts and can start to think independently about how to transform the organisation
We delivered a minimum viable product to traders involved in the initial process, which we now need to redesign based on learning in order to make it a more general product. We delivered a minimum viable company for spin out in line with ING’s Innovation and Platform strategy”
According to Santiago, the immediate plan for Katana Labs is to complete a successful funding round to support their business plan, which encompasses scaling up their client base and taking Katana from a discovery tool to a fully developed platform for relative value trading.
Below you can get a glimpse of the video we had created for Katana before the spinout.
