A silver lining in the cloud for AI startups

Cloud computing startups take pole position as more activities move online during coronavirus lockdowns

Data storage, processing and management are critical now as most activities worldwide move online during coronavirus lockdowns. Migration to the cloud was happening even before this crisis but now, organizations are depending on cloud services more than ever.

Whether it is working from home or streaming videos from Netflix, these services are underpinned by software and hardware on the cloud. Although there may be a reduction of cloud usage as airlines and others see business dwindle, the dominant theme in these times is digitalization.

The best-known cloud infrastructure providers are Amazon, Google and Microsoft. But there are several others catering to specialized needs. California-headquartered Fortune 500 company NetApp, for example, provides cloud services for storage and management of enterprise data. It has been in the thick of coping with a surge in demand. “Our leadership team is monitoring the situation daily to mitigate impact on services," says Ravi Chhabria, MD of NetApp India.

Tech innovations play a key role in cloud-based services ranging from telehealth to ordering groceries. “For example, AI can help ensure the supply of food to billions of people on lockdown around the globe," says Chhabria.

TAPPING INNOVATIONS

NetApp has been running a startup accelerator programme in Bengaluru since 2017 to tap innovations. These could enhance NetApp’s core technology. But the broader idea is to also take these innovations to NetApp’s customers.

“Our programme offers to help B2B tech product startups tap into the NetApp customer base globally, because we’re also a B2B (business-to-business) company," says Madhurima Agarwal, who leads the NetApp Excellerator in Bengaluru.

This is more significant in the current environment where travelling to meet prospective clients or vendors is not possible. The presence of facilitators like the NetApp accelerator can help overcome trust barriers and make connections between startups and enterprises.

The programme has evolved from its first cohort in 2017, even as the deep tech startup ecosystem in Bengaluru has matured. So all startups in the fifth cohort that graduated this year did a paid proof of concept (PoC) for NetApp, unlike the first batch which had very early stage startups. “Legal hassles and paperwork are taken care of at the PoC stage when startups are on-boarded as vendors. So the roadblocks from PoC to actual commercial engagement will be minimized," says Agarwal.

Bengaluru-based Lightwing, for example, automates the optimization of cloud services for enterprises, helping them cut cloud bills, which has become more vital in an economic slowdown. Lightwing’s PoC was to help business teams in NetApp optimize their cloud usage and measure savings.

“Then it becomes a matter of our sales teams going to our customers with Lightwing and saying, ‘This is how we’ve used it in NetApp’. We’re validating the product, and lending credibility and going to market," says Agarwal.

Eight startups from previous cohorts are in “strategic alliances" with NetApp for a go-to-market programme. Bengaluru-based Zscore, which uses AI to sort the wheat from the chaff in data, has engaged with NetApp clients in Australia. AI-powered cybersecurity startup Anlyz, which graduated in 2018, tested its product in NetApp last year before deploying it with a security service provider.

Corporate accelerator programmes have been a mixed bag in Bengaluru. Microsoft’s accelerator programme, for instance, has morphed time and again as concepts that sounded good on paper couldn’t be put into practice. Sales teams often have different agendas from those running accelerator programmes in large organizations.“It’s not easy," she admits. Sometimes it’s NetApp clients that move things along. “A client comes and says, ‘I saw this startup in your accelerator programme. What do you think of them?’ That motivates the sales team to come and talk to us," says Agarwal.

DRAWING ON CONNECTIONS

“We’re not humongous like an IBM or Microsoft, and we’re not small. We’re at that sweet spot where we still have a lot of interpersonal connects, and cross-functional teams can bank on one another," says Agarwal.

The idea behind the programme was to collaborate with some of the best minds outside NetApp that were working on cutting-edge technologies in adjacent areas of NetApp’s data business. “We look at them as partners and together we can create strong offerings for customers."

Bengaluru-based Curl Analytics, for example, is a “global alliance partner" for NetApp. It has a suite of enterprise software products: Sara is for intelligent data extraction from documents, Yati is a decisioning system, and Paras is an automated machine learning engine.

Curl Analytics deployed Paras in the AI CoE (centre of excellence) in NetApp’s Bengaluru campus. The CoE is equipped with Nvidia’s DGX box which has petaflop-level speed and scale for complex AI challenges. NetApp’s Ontap software is installed in it as an operating system. Curl Analytics’ solution demonstrates what can be done with mammoth datasets in such an environment. That makes it a win-win-win where all three companies showcase their strengths to enterprises, which see a live demo and visualize their own use case.

For the PoC in the accelerator programme, Curl Analytics deployed Paras on the Nvidia DGX and NetApp Ontap. The use case was to analyze millions of medical slides to figure out which ones had pathogens. “We showed how significantly the DGX machine, which is Nvidia’s supercomputer kind of thing for deep learning, can reduce time for training the AI engine," says Shivaram K R, co-founder and CEO of Curl Analytics. “We showed how this can be transferred seamlessly to cloud with NetApp’s Ontap and our Paras, so that doctors can draw inferences."

The next steps were to work out a go-to-market strategy where NetApp and Curl Analytics sell their products together because that’s when use cases become powerful, says Shivaram. “We are one of NetApp’s AI partners," says Shivaram. “For it to bear fruit takes time."

When the pandemic broke out, Curl Analytics and NetApp paused to take stock. The accelerator programme is slated to continue in virtual mode with everything from selection to mentoring and sales talks happening on video conferences.

“We changed our business model and increased our services-play to get cash flows," says Shivaram. “We asked large companies using our products if they want us to do services and they’re happy to offload their work. Normally we were downplaying our services, but now we are pushing it so that we have sustainable revenues. I believe we will come out strong once the dust settles down."