Traditional RPA — A Wasted Effort? Why Clear Ventures Invested in Kognitos

By Rajeev Madhavan, Co-Founder and Partner at Clear Ventures

Has robotic process automation (RPA) lived up to its hype? The RPA market has been booming as organizations are launching their digital transformation initiatives by using software robots or bots to automate routine and repetitive tasks such as data entry, data extraction, and invoice processing. RPA offers compelling benefits: it can boost productivity, efficiency, accuracy, security, customer service, just to name a few. With RPA, companies get to focus on what matters most, higher-value human work — the “bots” do the rest. The promise sounds great, but the reality is quite different.

Up to 30 to 50 percent of RPA projects fail, according to a study by Ernst & Young. And many fail on crucial deliverables: implementation time, cost savings, scalability, and the actual bot automation itself. These implementations often lack a clear strategy and vision and focus on automating specific tasks rather than the bigger picture.

RPA companies emerged from the consulting business as these companies were searching for ways to streamline business processes and reduce costs for their clients. They identified a clear need to use technology to automate repetitive and manual tasks, which led to the development of RPA. They essentially added a software script to automate certain routine, rules-based tasks, then added artificial intelligence (AI) to make the functionality easier, but these solutions are still fundamentally flawed. Consider this: for every dollar a company purchases in software, they may spend nearly 10X that amount to hire consultants/service providers to do the integration. [It is not uncommon for implementation costs to exceed software costs, especially for large complex systems] And when business logic changes — as it always does — they must invest even more in providers to do the automation.

Why do RPA projects fail?

Getting RPA projects right with current technology is difficult — and most companies simply underestimate the cost and complexities involved.

Ernst & Young cite reasons for failed RPA projects including not making RPA a business-led initiative vs IT-led; underestimating what happens after processes have been automated; treating robots as a series of automations vs. an end-to-end program; targeting the wrong processes or automating too much of a process; forgetting about IT infrastructure; and assuming RPA is all that is needed to achieve a great ROI.

But these projects also fail for the reasons below as well:

• The project requires too much complexity — many available RPA tools don’t handle complexity well, especially if there are too many steps required in the process. Forrester analyst Craig Le Clair recommends that traditional RPA be governed by the rule of fives: “No more than five decisions, no more than five apps, and no more than 500 clicks.” In other words, when you get beyond a certain level of decision complexity, the bots become very hard to maintain.
• The project fails to scale, which leads to isolated islands of automation with different departments using different bots, leading to inconsistencies and inefficiencies. Companies often get stuck after deploying only a few bots with up to 50 percent of RPA deployments initially failing. And bots face scaling issues when compared to API integrations — things like the complexity of processes being automated, number of applications and systems that the RPA bot must interact with are all factors.
• Other issues revolve around governance and RPA’s inability to deliver sufficient ROI; management issues once the digital workers are deployed to production; third-party issues because third-party interfaces aren’t uniform (bots are inflexible and don’t respond to variance); unrealistic expectations; shadow deployments; lack of buy-in from employees or resistance to the idea that bots are taking over their jobs — leading to lack of engagement/company culture issues; inability to handle change management, etc. Bots and consulting service teams are often seen as not aiding the business team or IT, but maintaining and managing them is seen as taking away resources.
• When problems occur or automation does not work as intended, the finger-pointing begins. And troubleshooting these issues in high-stress situations such as quarter closings can lead to rushed debugging and even more problems.

The market need

At Clear Ventures, we believe that RPA has not delivered on its promise. The market is ripe for a tool that will not just shift the problem from one department to another — say from finance to IT — but solve the enterprise automation issue and become more efficient with AI technology.

In looking at these challenges, Kognitos reasoned that instead of humans learning to use software, the software should learn to work with humans, giving humans the superpower needs to get their work done. Further, they asked; what if a solution could eliminate the issues that many enterprises today face with current RPA solutions?

• Remove the need for developers and services, thereby lowering total cost of ownership, enabling ever more complex processes to be automated with solid ROI.
• Remove the need for IT staff time to manage “bots.”
• Remove the need for learning yet-another new, complex tool.
• Leverage a tool that automates in plain English — natural language processing and large models applied to the core of the problem.
• Remove the fear that automation will fail — handle exceptions with intelligence like the best service providers can do.
• And allow automation of processes that traditionally were unfit to automate based on RPA best practices. Ultimately, be the Jarvis (Ironman assistant) for humans automating business processes!

Kognitos — addressing one of the biggest challenges businesses face today.

Enter Kognitos, founded by Binny Gill, who had an impressive eight-year run as a technology leader at Nutanix, the hyper-converged infrastructure company, eventually becoming its CTO for cloud services. Binny left Nutanix to start Kognitos based on a bold vision to make everyone a programmer. He created the company to address one of the biggest challenges businesses face today with automation — the lack of skilled personnel in automation. By enabling business users to describe even intricate processes such as invoice processing, insurance claims, credit card payment reconciliation, in English, Kognitos aims to radically accelerate the speed of innovation and time to value in businesses by reducing both the time and resources required to automate and maintain processes.

Kognitos is assembling a world-class team of enterprise sales leaders and process automation experts from leading customers or RPA solutions. The team is well-equipped to deliver on their promise of making everyone a programmer.

Further, Kognitos is working on enabling businesses to build their own personal “ChatGPT” for business process data that is private to the organization. In particular, the company’s solution will make processing data across multiple applications possible. This will ultimately unlock higher-level insights and over time improvements in business processes in a virtuous cycle of innovation and automation.

In early February, Clear Ventures led the oversubscribed $7.4 million seed round of financing for Kognitos, a pioneer in Generative AI for Automation using large language models (LLMs) such as GPT3 and ChatGPT, along with participation from Engineering Capital and Wipro Ventures, the corporate arm of Wipro. The company will use this funding to grow their marketing, engineering, and applied foundation model research functions.

“Current day automation tools require significant involvement of IT and service providers to use the no code/low code solutions. Kognitos has built a platform that allows the people who know and own business processes to combine business logic and Generative AI leading to rapid automation — all in English,” said Rajeev Madhavan, Co-Founder and Partner at Clear Ventures.

Kognitos is unlocking the power of AI for humanity with its generative AI platform — Koncierge. Says Binny, “Now every person will be able to use Generative AI for automating what they want — utilizing the English language. It’s time for computers to behave like humans, and humans to stop behaving like machines.”

Listen to Binny Gill discuss how LLMs like ChatGPT are making us all programmers, on the podcast AI and the Future of Work.

Learn more about Kognitos here and search for open positions in marketing, engineering, and in large language-model disciplines. The company is looking for UX/UI and staff engineers, and more.

About Rajeev Madhavan

Rajeev has been an angel investor since 1998 and co-founded Clear in 2016. Prior to that, Rajeev founded three successful startups in the semiconductor IP and Electronic Design Automation (EDA) space. The most recent, Magma Design Automation, became the fourth-largest EDA company in the world under his leadership.

Today, Clear Ventures has $335 million of capital from a world-class investor base and is an investor currently in about 25 active companies, 90% as first investors.