• Secures sensitive data with up to 50% increase in data loss prevention alerts and up to 25% reduction in leaks

• Enables System and Organization Control reports that validate platform's internal security controls to build client trust

• Frees up funds for platform development with $50,000 per year savings in collaboration tool costs

As Chief Risk Officer for a global bank, Pankaj Kulshreshtha wondered whether he could transform credit underwriting, the field of assessing loan applications, with artificial intelligence (AI). He believed that financial institutions were missing out on a big opportunity by relying on legacy methodologies that disregarded the massive proliferation of data available to better evaluate credit risk.

Pankaj launched Scienaptic on a mission to crack a seemingly intractable challenge. How to significantly increase loan approvals (boosting business), while simultaneously decreasing credit losses. His answer was to unleash the new wealth of nontraditional credit data, such as telecom bills and social media interaction, across AI neural networks to provide a more three-dimensional picture of an individual's credit profile.

For a small team like Pankaj's, solving a problem of such scale was a highly complex task. It requires seamless, real-time collaboration across the globe, between New York headquarters and data engineers in Bengaluru, India, as well as close communication with clients primarily in the US market.

Equally important is its world-class security protocols, to ensure no sensitive data leaks to outside parties in the process of solution development or client onboarding. This is particularly critical in the financial industry which is bound by strict data privacy protocols such as GLBA in the US, which requires financial institutions to develop a written information security plan to protect nonpublic personal information of clients.

With that in mind, Scienaptic turned to Google Workspace to enable and secure its global creative workflows. It discovered the seamless collaboration and robust security solutions it needed to build the credit underwriting platform of the future. Today its Ether platform enables up to 40% higher loan approvals with up to 25% fewer credit losses. In 2019 alone, Scienaptic platform enabled financial institutions to evaluate 45 million consumers and offer credit to more than 15 million people.

"Our mission to transform credit underwriting with AI required intense global collaboration and rock-solid data security," says Rakesh Agrawal, Product and Technology Leader, Scienaptic. "We simply could not have succeeded without the data loss prevention tools and productivity enhancements of Google Workspace."

In the quest to forge a new AI credit scoring paradigm, Pankaj and his team faced two key obstacles. The first was bringing products and updates to market fast enough to serve clients' real-time needs. The second was fulfilling "explainability" requirements in US equal credit opportunity laws. These regulations mandate that lenders be able to point to the exact data variable that causes a loan application to be rejected.

For years, banks had been designing AI models to predict creditworthiness. What limited the use of AI in credit scoring was the difficulty explaining exactly where in the web of AI neural networks, involving millions of interrelated data points, the algorithm decided to turn down a loan applicant. This was until Scienaptic came along. The startup deployed a data science technique called bucketing to slice scoring variables into discrete segments. This allows the platform to point to exactly which data factors led to the decision, satisfying regulatory requirements.

Scienaptic began developing its data bucketing solution using the Google Workspace basic plan to enable collaboration and productivity. The Google Workspace secure-by-design infrastructure gave the team substantial peace of mind, says Guru Prasad, Senior IT Manager, Scienaptic. But it soon ran into the need for more advanced, hands-on security tools and controls to safeguard the sensitive data and machine learning code that was criss-crossing the world every day.

In particular, Scienaptic sought a superior data loss prevention (DLP) solution to secure its AI enterprise. DLP safeguards data by alerting the platform administrator when sensitive information risks crossing the firewall, for example in an email, and in highest-risk situations automatically blocks the leak. In addition, Scienaptic needed an intuitive security dashboard to effectively marshal safety settings and automated alerts that protect information, identities, and devices.

After considering options, including rival productivity solutions, the startup decided to upgrade to Workspace Enterprise, Guru says, due to its versatility, intuitive interface, and superior asset protection tools. It found the premium solutions it needed to drive authentication, data security, and operational control, all from one easy-to-use dashboard.

On the Workspace Enterprise edition, Scienaptic gains unit-level data controls across the entire suite of Google Workspace products, including Drive, Google Meet and Gmail. Additional administrator privileges have been particularly important in giving Scienaptic a 360 degree view of security protocols, especially to carry out System and Organization Control (SOC) audits. SOC reporting provides validated intelligence on a service vendor's internal security controls to clients of the service, critical to building customer trust.

Since switching to Workspace Enterprise edition, Scienaptic has achieved significant results in bolstering the security of its internal systems and outside communications. In particular, it receives up to 50% more notifications of DLP threats, while reducing data shared with external parties by up to 25%. Guru estimates that DLP enhancement and premium administrator features, gained from the Workspace Enterprise upgrade, combine to give Scienaptic an overall 60 to 70% improvement in platform security.

"Upgrading to Workspace Enterprise transformed our ability to handle complex security tasks like data loss prevention and SOC reporting," says Guru. "It enables us to build trust with valued clients by demonstrating the safety of our systems."

Scienaptic's success depends on constantly evolving its AI solution to stay ahead of fast-paced financial industry changes. The productivity gains unleashed by Workspace continue to be instrumental to the critical task of bringing new products and features to market in a competitive environment, especially amid COVID-19 restrictions.

For example, the ability to receive real-time input from multiple global teams working on a single Sheets spreadsheet, with Workspace Enterprise level security controls, has yielded incalculable efficiency gains, says Guru. Meanwhile, Forms has proved critical in getting the entire worldwide team up to speed on internal trainings through interactive surveys and quizzes.

The ability to record, store, and easily search collaborative sessions, such as Meet conferences, using Cloud Search has enabled Scienaptic to forgo a separate business communication platform it had been using to organize information. That is saving it an average of $50,000 per year in costs, which it is then able to plough into application development.

"With Meet, Drive and Cloud Search, we've been able to accelerate our solution during the pandemic," says Rakesh. "In fact we've been able to boost productivity through remote collaboration, rather than decrease it, and that's been amazing."