The emerging trend in healthcare is the use of data analytics, which can predict, for instance, disease outbreaks and, at an individual level, whether a person is likely to develop a certain illness, or a woman is likely to deliver a baby prematurely.
In November 2023, the UK Biobank released the full genetic sequences of half a million people — a treasure trove that scientists could use to uncover links between DNA and diseases.
By analysing these genomes alongside clinical data, researchers can identify genetic markers associated with diseases, predict patient outcomes, and tailor treatments based on individual genetic profiles.
This approach holds immense potential in tackling complex conditions such as cancer, cardiovascular diseases, and rare genetic disorders.
What UK Biobank released was a large dataset, but it is truly tiny compared to the ocean of data lying out there. Billions of pieces of data, including images (X-rays and other scans), are awaiting the attention of researchers.
Incidentally, researchers at AIIMS, Delhi, are collecting images of oral cancer and pre-cancer cases and labelling them; 5,000 labelled images in hand, they aim to do ten times as much.
A hospital chain in the US is using data to analyse the pathology of kidney donors and recipients to determine the right match.
Similarly, there are mountains of data on the side of the commercialisation of drugs.
Data abounds for deciphering everything from patient behaviour management to how a drug sales representative can best approach a particular doctor.
But data exists in silos and doesn’t readily yield to analytics (picking up non-obvious trends). Someone needs to ‘integrate’ the data to make it ‘analytics-ready’.
A US-headquartered start-up called Agilisium Consulting, founded by Chennai-origin Raj Babu, does the job of getting data analytics-ready.
Agilisium, in a way, encapsulates two entwining trends in the data industry — the business opportunity in integrating and ‘cleaning’ data to make it suitable for analytics; and the drastic reduction in cost, thanks to cloud computing.
Babu earlier worked for 20th Century Fox and Universal Music Group, where he had to rummage through data and suggest the best times to display or withdraw a movie, or place a DVD on a Walmart shelf.
Having cut his teeth in the field, he combined the ‘agility’ of data with his favourite Matt Damon-starrer sci-fi Elysium to start up Agilisium Consulting a decade ago. Its services include data architecture consulting, data integration, and storage and analytics.
In a way, Agilisium is like ChatGPT at an enterprise level. Over 70 per cent of its 850-odd staff is stationed in Chennai.
In a conversation with Quantum, Babu gave the example of Amgen, a US-based biotech research company, which was overwhelmed by the amount of data sets it had and sought Agilisium’s help to handle it.
Asked if managing so much data requires large computing power, Babu says the company hires compute power from the cloud.
Agilisium is an ‘AWS or Amazon Web Services partner’, which means it provides AWS cloud services and solutions to customers.
Amgen says its data processing time dropped from 48 to 12 hours and it was able to ingest several petabytes of new data sets without disrupting the system performance.
“We grew from hundreds to thousands of users, while reducing the errors in our data metrics,” says Sheetal Pillai, Amgen’s senior manager in charge of commercial data sciences. These new approaches to data management and processing enabled Amgen to scale up quickly, reduce time-to-market, and unlock more computing power.
Agilisium recently announced that it has set apart a million dollars for co-innovation.
Its pitch: If you have a problem and the data that can be used to solve it, then it will invest $25,000 (in terms of professional time, buying cloud capacity, and so on) to provide a solution. Agilisium’s takeaway will be the learnings as well as a chunk of codes, which can be applied elsewhere.
None of this would be possible without the tremendous fall in the cost of hiring cloud services. Today, you can store 100 TB of data on Azure for $1,600 a year. Amazon S3 offers the same for $2,100. Earlier, when you had to buy your server, it cost millions.
Babu says India can become the “data-driven digital R&D hub of the world”.