SupportLogic: Leveraging Machine Learning to Transform Customer Support

Krishnaraj Raja, Founder & CEO

Krishna Raj Raja stepped into the realm of customer support when he became the first hire in India for VMware (he founded the VMware Bangalore office in 2004) and their fifth technical support engineer. Krishna ascertains that support engineers are the real product experts and that they have a lot of valuable knowledge of customers, their needs, concerns, and habits.

We drive digital transformation by extracting latent signals from support conversations

Tribal knowledge is extremely useful not just for support but also for teams rooted in sales, product, engineering, and marketing departments. However, this tribal knowledge is harnessed very poorly by organizations today. First of all, extracting tribal knowledge is very hard and all that is available today for organizations is metadata that is captured as part of a support case. Using metadata to describe this knowledge is like using the names of colors to describe Mona Lisa painting. Secondly, the metadata captured in support is often inaccurate. Almost all organizations struggle with data quality issues (classic garbage-in-garbage-out problem). The real knowledge that is there on the support engineers’ brain is lost when the person leaves the company. Krishna’s says that this was his experience at VMware.

The simple fact is that within any organization, many consumers feel that their complaints either go unheard or take a long time to be acknowledged and resolved, which can lead to high levels of discontent. This affects the long-term outcome of the company.

So, what is missing? What’s going wrong? To begin with, ticketing systems are not capable of storing, analyzing and contemplating customer interactions. They are mere databases, they lack cognitive abilities, and they are not content or context-aware. Today, we assume that tribal knowledge can be gained only by reading every incoming ticket, but this doesn't scale as the ticket volume increases. Ticketing systems today capture a tiny fraction of the actual customer conversation, which is not sufficient to service them efficiently.
Sometimes, even these small pieces of data are lost as service agents attempt to translate them into actionable measures. Despite the investments that companies unleash on the systems of records, they fail to reap the benefits due to the lack of a mechanism to holistically address their customer needs or gain (capture) any insights from the interactions with them. This hampers their growth, brand value and, most importantly, revenue.

Having worked on the customer front for organizations such as VMware and CloudPhysics, Krishnaraj Raja witnessed and has even faced these predicaments first hand. He realized that organizations often lose out on a lot of vital information or “knowledge” about the customers, as they had no way to efficiently store the data and leverage it to improve their services. Additionally, many organizations adopt multiple systems of record that serve the same purpose for different teams, such as support, sales, IT — bringing about data silos, and hindering proactive collaboration and collective decision making.

As there wasn’t a unified solution to address these issues, Krishna decided to take it upon himself to help organizations improve their services and support functionalities by providing them with a solution that can enhance communications, store data and pave the way for firms to intuitively access and leverage customer information at the right time.

Enter, SupportLogic! On a mission to transform the role of technical support delivery, this California-based Information Technology and Services startup offers a machine-learning-based system of intelligence that extracts and distills information from tickets, understand its urgency and context, and even provides critical insights for organizations to scale customer interaction and orchestrate actionable decisions.

SupportLogic’s easy-to-deploy platform seamlessly integrates to their clients’ existing systems of record through APIs and an in-house cloud, which enables them to quickly get up-and-running in weeks without the need for any IT or computer resources or investing time or money in replacing tools. It identifies and extracts signals or key information from customer interactions using Natural Language Processing and Neural Networks. It also applies ML to gather critical insights that firms can leverage to serve their consumers better, providing business intelligence, workflow automation, real-time alerts, and predictive services.

“We use NLP and Deep Neural Networks to extract latent signals and tribal knowledge residing in technical support conversations,” says Krishna.
“We help our customers scale their interactions with customers (internal or external) by notifying them at first indication that something might be going wrong.”

SuportLogic uses the signals that it extracts to provide an early warning system to prevent customer escalations, also enabling them to predict escalations and identify the factors that are driving them. Not only does this guarantee customer satisfaction, but it also paves the way for firms to better their product or services. SupportLogic also focuses on increasing collaboration between teams within organizations and allowing them to share the tribal knowledge. SupportLogic offers an easily accessible self-service dashboard on which users can observe all the interactions, learn from the gathered information and intelligence, and together, can act proactively.

The success story of one of SupportLogic’s clients best exemplifies the company’s capabilities. Krishna highlights an instance wherein, an organization growing at a rapid pace witnessed an increasing number of tickets pouring into their systems of record daily. Consequently, they were unable to help customers effectively and risked losing them. The sudden surge in support requests put customer-facing teams in an awkward position, as they could not invest adequate time in analyzing the tickets, prioritize them and take relevant actions. The firm was unable to solve their clients’ problems effectively, which lowered customer satisfaction dramatically, which also hampered brand image and revenue.

After deploying SupportLogic’s offering and leveraging the early warning system embedded within the tool, the firm was able to focus on consumers based on the system’s prediction and bring in the right people to resolve tickets, which made all the difference. The company was soon able to address problems faster, increase customer satisfaction and scale customer interactions.

In the span of merely three years, SupportLogic, with founder Krishnaraj Raja at the helm, has helped many such organizations and has onboarded serial investors such as Tim Guleri from Sierra Ventures to serve its clients better. While the company works mainly behind the scenes, the product has been deployed in production and used by some of the biggest names in the IT industry.

“We want to disrupt this industry by helping our clients focus on understanding and solving customer issues, rather than deflecting it—which not only ensures a higher customer satisfaction rate but also paves the way to better a product or a service,” concludes Krishna.