Hello, Rodrigo. Thank you for joining us today.
Please tell us a bit about yourself and your current role with Aridhia.
Sure. So, a little bit about my background.
I'm a software engineer and kind of a bit of a mathematical background.
I was one of the first employees at Aridhia and
I've always worked on product and product strategy.
Working at Aridhia has given me the opportunity to work
with research hospitals across the UK,
and also a little bit further afield like Kuwait and Australia.
My main technical interest is in making healthcare data more productive,
and finding ways for it to drive innovation within healthcare systems.
Please tell us a bit about Aridhia's platform.
What does it offer? And why is it needed?
It's an online analytics environment for collaborative healthcare research,
and other kind of analytics usage.
And it came out of a collaboration between Edinburgh University,
Dundee University, and the cancer centers within NHS Tayside and NHS Lothian.
And our goal really was to, initially with Cancer Informatics,
but then more generally with Medical Informatics.
How do you make more data available,
and then create an environment which is secure and safe, for research to be undertaken?
But also to do that in a way that can scale,
so not just at an individual project that can get us that kind of environment,
but anybody can sign up and hospitals can get these collaborations up and running.
And so we have two parts of the platform at a very high level,
one part is the Collaborative Workspace Environment, where scientists, clinicians,
industry, technical specialists can come together and do the kind of analysis,
particularly on privileged data,
patient data in particular.
And the other part is, we're currently calling it a Healthcare Landing Zone,
which is a kind of bridge between the hospital and that collaborative environment,
so that hospitals can safely make data available,
let researchers select the data,
get approval on that data usage,
and then carry on with their project.
So, our goal was really to provide a commercial service that people could rely on.
And through that, we can streamline access to data,
and we can support a kind of range of everything from kind of classical, statistical,
medical statistics, to the larger analysis
that you need with genomic datasets for example.
And we do that to a level of service and security that people can contract.
So what tools are integrated into the platform?
So, our main approach has been to,
from an end user point of view,
provide open source analytical tools.
And that's really driven by the diversity of our users,
the type of projects they want to do, their backgrounds,
and their desire to develop analysis in their project,
and then take it to the next project and build on it over time.
And an open source really helps them with that.
But we also help people bring proprietary tools to the platform.
So, we do that in a couple of ways,
and we know that our customers like universities or hospitals,
they all have their own tools that they've
developed in-house that they would really like to use.
And also they're in partnerships with commercial providers,
and also they just want to use proprietary tools themselves,
and so we offer the ability to deploy those.
Users can install their own software in a kind of virtual desktop environment.
It's like a familiar environment that they would have had in their own machine,
but it's online and it's within that kind of secure environment.
So what do your customers look for from the security point of view?
So a lot of our customers, being hospitals or research hospitals,
are data owners or data controllers.
And so, they know they need to do more with data,
particularly clinical data, hospital data.
But they're held back by security and governance constraints.
Our background is the Scottish Health Informatics Program which
really set out some good principles and guidelines,
best practice around how that relationship between data owners and data users could work.
And so we've been applying that in the environment.
And as one of our customers in the Netherlands says,
our workspace provides a legal space for the research to happen.
So it creates the right conditions for the research to happen.
And what does that mean in practice?
In practice, it's a series of
technical and operational things we have to do as a service provider.
We have to provide the same level of service on
a shared platform for lots of different projects
at the same time, we call that multi-tendency.
We have to help the customers meet
their own legal obligations, by using features of the platform.
So, we have to achieve certain accreditations.
We're tested by a third party, so called penetration testing.
So, security consultants would come in and see if they can break our system.
We have to be doing that four or five times a year, just to keep up with the threats,
the cyber security threats that are out there.
And at the same time,
we also have to make sure our processors and
our software development is of a very high standard and a secure process in place.
How does that impact users to do their research?
So for most of our users who are very,
very focused on their research goals,
they're writing a PhD or a paper,
that's really paramount or they want to improve something in their clinic,
that's the most important thing.
And we try to provide that legal space,
that secure space without impinging on their work.
So, a comprehensive audit happens in the background.
They don't need to worry about
recording a lot of what they do, it's happening automatically.
The environments are familiar,
there are a familiar web interface source or secure remote desktop.
So, we try to get out of the way from the point of view of providing lots of barriers,
but we do have to give certain controls in terms of
how we make sure that the right people are logging in to the right services,
and so we put in place those technical measures.
In general, we want to make our environment appealing to users,
so it can't be too locked down.
And we understand most of our users these days travel a lot,
they need to be able to work on their research on the go.
That's the modern way of life and research service. So we work around that.
What collaborative work is Aridhia involved in?
Collaboration has been key for Aridhia to get our platform off the ground,
to develop it, to enrich it.
And these days, we're more focused on helping other people collaborate.
But we are involved in, for example,
European-wide IMI project, between
universities and pharmaceutical companies around Alzheimer's dementia.
An IMI project is a collaboration between the European central funding,
the governments, and the International Pharmaceutical Industry.
So it's a bit like Horizon 2020.
It's a big European grants that are available to universities, SMEs, and hospitals.
And similarly with Medical Research Council or
the Chief Scientist office here in Scotland,
we're involved in a number of collaborative projects,
and the key to those is bringing people to
some common ground where they can work together on the platform.
We're particularly interested where a lot of our customers are asking
us to create hybrid systems between the clinic and research.
And to do that, they have to collaborate,
they have to break boundaries, in order to just achieve that.
So, helping them collaborate is so important.
How is Aridhia increasing data availability for researchers?
Data availability, as I think I said earlier,
is one of the important needs of our research hospital customers.
And it's a really good area of development at the moment,
and we've been working with,
for example, Great Ormond Street Hospital
and Radboud University Medical Center in the Netherlands,
who are kind of pioneering this field.
A lot of our community use the principles of FAIR.
So that's an acronym that stands for data should be findable,
it should be available, which means you can get to it.
It's interoperable which means that the meaning of it is understood and you
could combine it with other data to broaden your analysis.
And it's reproducible, in the sense that you can
re-run your analysis and get similar results and validate other people studies,
and other people can validate your studies.
And so we try to build that into our system,
the way we provide the service.
So, users can find data.
We can help them by describing their data in a catalogue.
We can also help them document the detail of the data at a field level through a dataset,
what we call a dataset library service,
which allows other people to find it.
Someone has data similar to what I'm looking for, for my project,
and some of these concepts of metadata are quite complicated for users,
so we try to make our services more user-friendly and focused on quite clear steps.
So for example, I want to translate my data from
the codes that the hospital uses to the codes that are used in research.
So, you might have heard of SNOMED.
It's a medical vocabulary that's used by hospital systems.
It may not be used a lot by researchers.
Some researchers use a coding scheme called Human Phenotype Ontology.
And so, we provide a service to help them translate from one to the other.
At the moment, we're working in an area of making the ability for users to
define the data they need and what data is
available much more streamlined and self-service process.
A lot of the process of finding data
is slowed down because of the concerns about governance.
But it's also the systems haven't developed so much,
because people are just afraid of moving the data forward.
So, working with the institutions and hospitals,
we're trying to see what's
an appropriate level of automation and self-service that can be put in place.
And that's already accelerating the number of projects that can get
data from this great resources that exist within the hospital systems.
How do your users progress as data scientists within the platform?
So, it's very important to keep our users happy and productive in the environment,
and we know that most people in this field are constantly learning new tools.
So, our users can start working from a file, a CSV file or something,
and have it de-identified and work at that level using tools that they're familiar with,
for example, in a Virtual Desktop.
But then they can progress to storing more data in a different way,
maybe using a big parallel database running in Database Analytics.
They maybe want to start using more programming paradigms, for example,
R or Python and start developing scripts and tools,
use third party libraries and packages for that.
And as they develop that,
they maybe want to scale across more data,
or more intensive computations.
And so we're kind of we're there for them by giving them access
to a more on-demand computing of different types.
So they can really run more,
that more complex training.
For example, if they move from a Statistical Analysis or
a Linear Model to more complex Machine Learning,
they can step up through that kind of spectrum all the way from, as I said,
from something like a Linear Model through to Machine Learning,
even to modern techniques like Deep Learning.
What are the next steps for Aridhia's platform?
So the next steps for our platform are driven by a roadmap that's
our customers influence and we have a vision of how we'd like to deliver this platform,
and we update every quarter or at least once a quarter.
So there's lots of changes all the time.
But particularly, over the next year,
we're looking at scaling up what people can do in the platform.
So that's both scaling up the size of data they can work with,
the amount of computation they can run in the platform,
but also the diversity of tools that they can run,
and making more and more of that self-service.
You'll see this constant theme.
Researchers are very autonomous people,
and we want to help them with that by giving them the power to kind of run their systems,
so we'll follow that track a little bit more.
Okay. And how will this improve data analysis?
You're asking about improving data analysis itself,
and I think that's really important for the individual researcher.
But for the broader stakeholders,
I think there's a bigger picture that we're trying to improve.
So we're trying to help people
create these hybrid systems between the clinic and research.
So what's the innovation pathway between, hospital makes data available,
researcher does their analysis,
a new tool, the Decision Support Tool gets developed,
how does that get applied, tested, validated?
And then eventually, hopefully,
some kind of spin out takes it to market.
And so, there are well trodden paths for pharmaceutical development for example.
But I'm really interested in
informatics intensive innovation that's driven by the availability of this data.
Thanks very much for talking to us today, Rodrigo.
You are welcome.