The startup and entrepreneurial culture of Boston was the focus for my last two days in the US. This was an opportunity to hear about how the innovation thinking that I learnt about at MIT in week one has been put into practice in the private sector (see my previous blog).
Data science and data scientists have become hot property in the US and world-wide over the past year, as organisations try to unlock the valued their data assets.
Thanks to my New Zealand Data Futures Forum colleague Josh Feast, I was able to visit two data-driven startups ? Cogito and Tamr ? and meet with the Data Innovation Lab team at Thomson Reuters.
Datafied innovation for better health and better business
Cogito is the company that Josh cofounded seven years ago with Alex Pentland, a Professor at MIT and serial entrepreneur, who is one of the most cited authors in computer science. Josh is an expat kiwi, who started out his career at Child, Youth and Family, after graduating in computer science from Massey University. He won a Fulbright to undertake his MBA at MIT, and then stayed on. In August 2014, Josh's company was named one of the 10 Fastest-Growing Companies in Boston by Inc.com, with a remarkable 3-Year Growth Rate of 1,529%. It now has 35-40 staff spread across its customer services, development and research teams.
Cogito grew out of Josh's research at the MIT Media Lab on behavioural analytics. A key strength is using data driven technology for health services ? improving patient insight, facilitating research, diagnosis, treatment and monitoring. This is achieved by patients installing a smartphone app that tracks phone calls and voice streams, and provides smartphone trace data for analysis. In defence- funded research, this technology has been used to detect psychological issues and monitor risk of suicide or violence. Given the sensitivity of this application, Cogito has been careful to build in sound ethics and privacy by design from inception, in the spirit of Alex Pentland's New Deal on Data.
From his public sector social services experience, Josh has carried a strong social conscience and principles into the leadership of his company. This is demonstrated in Cogito's strong focus on research. I met with Skyler Place, Vice President, who leads the research and development team, who told me about the mental health research projects that Cogito is engaged in with health sector partners, and emphasised the need for strong science and validation in the development of tools.
Cogito has also deployed voice analysis technology to enable real time analysis and improvement of call centre customer interactions, in their commercial tool Cogito Dialog ? a strong area of growth for the company.
While Cogito's business model is based on the datification of everyday activity through smartphone traces, data-driven innovation in other organisations is often based on maximising the potential of existing data assets.
Transforming data assets into business value
Thomson Reuters is a knowledge business. The company holds a large quantity of data in well curated, trusted formats. The enterprise strategy is that growth of data presents strong opportunities to drive value at scale for the company of 54,000 staff.
Thomson Reuters established its Data Innovation Lab in mid-2014 to enable the organisation to develop new potential from its existing data. Thomson Reuters' data assets include vast international datasets on patents, science citations, genomics and biomarkers, as well as news media data.
The Lab, based in the Boston office, was created to enable customers and staff in other divisions to find the signal in the noise of big data" (in the words of Nate Silver). It is a lab in the sense that it applies the innovation processes of design and rapid prototyping to research potential applications, which may then be productised by the company.
I met with Mona Vernon, Vice President, Data Innovation Lab and her team of eight data scientists, as well as with Emmanuel Thiveaud, Vice President, Research Data Services and Market Development in Thomson Reuters' Intellectual Property and Science Division. They shared a number of examples of their work with me, including visualisations of trends, top rankings and lead times for twelve industries' inventions via patents, and a geopolitical risk dashboard for the financial division predicting levels of risk for specific companies in different counties based on publicly available information.
Mona's philosophy is that "you have to be deliberate with innovation". There are many great things you can do. You need to have the discipline to turn down requests for minor enhancements and instead focus on new developments, and to only pick up work that has a business partner with a great hypothesis who brings content expertise, commitment and feedback from customers.
Data science is plumbing as well as poetry
To do this type of work effectively, data must be well curated ? cleaned, indexed with metadata and standardised. Tamr is a Boston startup that is filling this need by replacing manual processing with machine learning. The company was founded in 2013 by database industry veterans Andy Palmer, Mike Stonebraker and Ihab Ilyas, following the success of initial research at MIT CSAIL. Its focus has been on building a commercial-grade solution to tackle the challenge of integrating and enriching diverse data at scale, quickly and cost effectively.
I met with Min Xiao and Alan Wagner to talk about how the startup is developing. Like Cogito, Tamr is growing fast ? and facing the startup challenge of managing its growth curve so that the company is not overstretched in its early years. It has quickly found itself in the frontline the big data movement with demand from a variety of sectors including pharmaceutical, supply chain, customer integration, government, and knowledge industries like Thomson Reuters.
Tamr has a connection to New Zealand ? Andy is a massive rugby fan and a 1990s All Blacks top signed by Jonah Lomu takes pride of place on their wall.