Who is Ioannis Tsamardinos?
Dr. Ioannis Tsamardinos is a Professor of Artificial Intelligence, founder, and CEO at JADBio. He holds a Ph.D. from Pittsburgh University, US, and has served as Professor at the Department of Biomedical Informatics, at Vanderbilt University. Ioannis’ research includes Artificial Intelligence Machine learning and Automated Machine Learning. Ioannis has several prestigious awards under his belt, including the ERC grant (the EU equivalent of the NSF’s Young Investigator Award) and a group award for Software of the Year for his participation in the development of the Remote Agent software while earning his M.Sc. and Ph.D., in collaboration with NASA researchers. One of Ioannis’ algorithms is on board the Deep Space I spacecraft. He is currently a faculty member at the Computer Science Department at the University of Crete where he heads his own Lab called MensXMachina, or “mind from the machine”, successfully translating academic R&D into real-world business tools ready to market.
What’s the story behind JADBio?
Founded in 2019, JADBio, is based on the cumulative knowledge and expertise —in Predictive Analytics and Causal Analysis— of its founders, Ioannis Tsamardinos, Vincenzo Lagani, and Pavlos Charonyktakis. In 2013 we created Gnosis Data Analysis, a University of Crete spin-off located in Heraklion, Crete, Greece to develop algorithms and machine learning solutions for biomedical data.
As the team grew, and significant funding started coming in, we embarked on building JADBio, a purpose-built platform that would provide leading-edge AI tools and automation capabilities enabling life-science professionals to build and deploy accurate and explainable predictive models with speed and ease, even if they have no data science expertise.
JADBio is a robust AutoML platform focused on biomedical and multi-omics data. We make it easy and affordable for data scientists and life-science professionals to use data science to discover knowledge, while reducing time and effort by combining an end-to-end machine learning platform for data analysts with a wealth of capabilities ranging from smart feature selection to the reuse of predictive models. By providing off-the-shelf models for common healthcare use cases and automating many of the manual processes involved in traditional data-science tasks, JADBio allows data analysts to focus on what really matters, the problem and its solution. JADBio is headquartered in the Technology and Science Park in Crete, Greece and has a branch in LA, California, US.
What was the most difficult part of your experience in the early beginnings?
It was probably the first 100 days, before launching JADBio, when we were working feverishly on releasing the product, running tests, collecting feedback. I can only describe that period as absolute out-of-control productive chaos. We were trying to make new hires, clarify and implement our strategy and objectives, define roles and responsibilities; establish budgets and build the foundation for scaling up our product. Furthermore, with the first contracts we had to be laser focused on balancing all the moving parts and meeting our service responsibilities. Basically, you throw a bunch of balls in the air and start juggling. It’s do or die really.
What are you most proud of regarding your business?
We designed this Automated Machine learning tool to empower anyone to turn raw biological data to actionable knowledge. I feel very proud when I see our non-expert analyst users, like doctors, biologists, pharmacologists, psychologists, and other life scientists perform a sophisticated – and correct – ML analysis using JADBio, publish the results, and put them into practice. I feel we opened for them a window to a new world. They have direct access to the knowledge in their data. A few years back, this would have required them to get funding to hire a team of analysts, wait a few months, and then get results.
But, I particularly feel proud when I see bioinformaticians and data analysts praise JADBio; even though experts could replicate the automated analyses of JADBio, they still find many aspects of it fascinating: the ease of use, improvements in productivity, in the way we estimate performances, visualize results, optimize clinical thresholds and numerous other details that we worked out.
Customers want to learn the patterns in their data to allow them to predict the future (e.g., who’s going to get a severe reaction to covid-19, which therapy will work best, etc.); they also want to discover the quantities involved that carry the predictive information (e.g., gene that determines the response to therapy) and discover new knowledge. Eventually, the goal is to discover new drugs, optimize therapies, optimize diagnosis, optimize screening, lifestyle advice, discover new biology and medicine, and improve human health. With data volume increasing, exponentially, and not enough expert time to hire …only 0.5% of the data is analyzed.
What is your vision for the future of JADBio?
Let’s face it, we are not short of ambition. We are striving to have JADBio democratize machine learning for citizen data scientists, the way Excel democratized spreadsheets. By bringing AutoML to lifesciences, the goal is to discover new drugs, optimize therapies, optimize diagnosis, optimize screening, lifestyle advice, discover new biology and medicine, and improve human health. Our vision is to save lives. On the technical side, we envision JADBio to encompass an eveen larger part of machine learning, including causal analytics that is key in lifesciences. We want to grow it to become the leading AutoML platform in this vertical.
What’s your advice for the businesses that are trying to adapt to this economic climate?
Research, learn, adapt! The client purchasing patterns change, the way people meet, socialize, and network, and the way deals are sealed. Investor patterns also change. A small business has to be agile, continuously monitor the situations, and be ready to adapt. Though the coronavirus has led to unprecedented disruptions for small businesses, the pandemic, although very serious, is a temporary problem. So, during such testing times, startup entrepreneurs will have to adapt and survive the slowdown; when the world returns to some sort of normality, they have to adapt again. But, for the AutoML industry, the future seems bright. The global market for automated machine learning, measured by revenue, was 270m USD by 2019 and the expected CARG for the coming 10 years is 44%, which is predicted to lead to a market of 14bn USD by 2030. Life-sciences are expected to be among the top adopters of this technology, with an obvious near-term high return on investment rates, and an immediate access to vast amounts of unanalyzed data that holds huge potential to be unlocked.
Please name a few technologies which have the greatest impact on your business
Obviously, machine learning developments are crucial, but as any SaaS platform out there, we are affected by storage, CPU, 3rd party SaaS vendors like Amazon’s AWS, even payment services like Stripe. Platforms are not developed in a linear line. While the platform is our main concern and we are adding new capabilities, like the analysis of very large feature sets, medical images and voice as data sources, in parallel we are mindful of creating a seamless customer experience for our users with the aforementioned technologies.
What books do you have on your nightstand?
I am a geek, truth to be told. “Convex Optimization by Stephen Boyd and Lieven Vandenberhe, Cambridge Press” for some light mathematical reading,“
Our mathematical Universe by Max Tegmark, Knopf”, a life-changing book about how the world works, and the “Cyberiad by Stanislav Lem” for entertainment.
Because of the current economic climate our publication has started a series of discussions with professional individuals meant to engage our readers with relevant companies and their representatives in order to discuss their involvement, what challenges they have had in the past and what they are looking forward to in the future. This sequence aims to present a series of experiences, recent developments, changes and downsides in terms of their business areas, as well as their goals, values, career history, the high-impact success outcomes and achievements.